Telematics Communication Technologies and Vehicular Networks: Wireless Architectures and Applications Chung-Ming Huang National Cheng Kung University, Tainan, Taiwan, R.O.C. Yuh-Shyan Chen National Taipei University, Taipei, Taiwan, R.O.C. InformatIon scIence reference Hershey • New York Director of Editorial Content: Senior Managing Editor: Assistant Managing Editor: Publishing Assistant: Typesetter: Cover Design: Printed at: Kristin Klinger Jamie Snavely Michael Brehm Sean Woznicki Michael Brehm, Carole Coulson Lisa Tosheff Yurchak Printing Inc. Published in the United States of America by Information Science Reference (an imprint of IGI Global) 701 E. Chocolate Avenue Hershey PA 17033 Tel: 717-533-8845 Fax: 717-533-8661 E-mail: [email protected] Web site: http://www.igi-global.com/reference Copyright © 2010 by IGI Global. All rights reserved. No part of this publication may be reproduced, stored or distributed in any form or by any means, electronic or mechanical, including photocopying, without written permission from the publisher. Product or company names used in this set are for identification purposes only. Inclusion of the names of the products or companies does not indicate a claim of ownership by IGI Global of the trademark or registered trademark. Library of Congress Cataloging-in-Publication Data Telematics communication technologies and vehicular networks : wireless architectures and applications / Chung-Ming Huang and Yuh-Shyan Chen, editors. p. cm. Summary: "This book examines critical issues involved with telematics such as vehicular network infrastructure, vehicular network communication protocols, and vehicular services and applications"--Provided by publisher. Includes bibliographical references and index. ISBN 978-1-60566-840-6 (hardcover) -- ISBN 978-1-60566-841-3 (ebook) 1. Automotive telematics. I. Huang, ChungMing, 1961- II. Chen, Yuh-Shyan. TL272.58.T45 2010 629.2'77--dc22 2009034903 British Cataloguing in Publication Data A Cataloguing in Publication record for this book is available from the British Library. All work contributed to this book is new, previously-unpublished material. The views expressed in this book are those of the authors, but not necessarily of the publisher. List of Reviewers Chih-Yung Chang, Tamkang University, Taipei, Taiwan, R.O.C Ruay-Shiung Chang, National Dong Hwa University, Hualien, Taiwan, R.O.C. Tzung-Shi Chen, National University of Tainan, Tainan, Taiwan, R.O.C. Wei-Kuo Chiang, National Chung Cheng University, Chia-Yi, Taiwan, R.O.C. Li-Der Chou, National Central University, Chungli, Toayuan, Taiwan, R.O.C. Chyi-Ren Dow, Feng Chia University, Taichung, Taiwan, R.O.C. Ren-Hung Hwang, National Chung Cheng University, Chia-Yi, Taiwan, R.O.C. Guan-Ling Lee, National Dong Hwa University, Hualien, Taiwan, R.O.C. Wan-Jiun Liao, National Taiwan University, Taipei, Taiwan, R.O.C. Tsung-Nan Lin, National Taiwan University, Taipei, Taiwan, R.O.C. Jen-Yi Pan, National Chung Cheng University, Chia-Yi, Taiwan, R.O.C. Shiao-Li Tsao, National Chiao Tung University, Hsinchu, Taiwan, R.O.C. Table of Contents Foreword ............................................................................................................................................ xix Preface ................................................................................................................................................ xxi Acknowledgment ............................................................................................................................... xxv Section 1 Introduction of Vehicular Networks and Intelligent Transporation Systems Chapter 1 Introduction of Vehicular Network Architectures ................................................................................... 1 Ming-Chiao Chen, National Taitung University, Taitung, Taiwan, R.O.C. Teng-Wen Chang, National Taiwan University, Taipei, Taiwan, R.O.C. Chapter 2 Introduction of Vehicular Network Applications .................................................................................. 15 Yao-Chung Chang, National Taitung University, Taiwan, R.O.C. Chapter 3 Introduction to ITS and NTCIP ............................................................................................................ 32 Da-Jie Lin, Feng Chia University, Taiwan, R.O.C. Chyi-Ren Dow, Feng Chia University, Taiwan, R.O.C. Section 2 Embedded System Architecture and Communication Protocols Chapter 4 Vehicular Embedded System Architecture............................................................................................ 58 Chung-Ping Young, National Cheng Kung University, Taiwan, R.O.C. Chapter 5 Data Communications Inside Vehicular Environments ........................................................................ 74 Cheng-Min Lin, Nan Kai University of Technology, Taiwan, R.O.C. Tzong-Jye Liu, Feng Chia University, Taiwan, R.O.C. Chapter 6 Wireless Access in Vehicular Environments ......................................................................................... 90 Tzong-Jye Liu, Feng Chia University, Taiwan, R.O.C. Ching-Wen Chen, Feng Chia University, Taiwan, R.O.C. Section 3 Location Based Services Chapter 7 Introduction To Global Satellite Positioning System (GPS) ............................................................... 108 Jenq-Muh Hsu, National Chiayi University, Chiayi, Taiwan, R.O.C. Chapter 8 Vehicle Location and Navigation Systems ......................................................................................... 119 Ben-Jye Chang, National Yunlin University of Science and Technology, Yunlin, Taiwan, R.O.C. Chapter 9 Design and Implementation of Vehicle Navigation Systems .............................................................. 131 Min-Xiou Chen, National Dong-Hwa University, Hualien, Taiwan, R.O.C. Section 4 Integrated Vehicular Application Chapter 10 Vehicular Metropolitan Area Network Systems Architecture: The WiMAX Network Reference Model ................................................................................................................................. 144 Cheng Hsuan Cho, National Chung Cheng University, Taiwan, R.O.C. Jen-Yi Pan, National Chung Cheng University, Taiwan, R.O.C. Chapter 11 Interworking of IP Multimedia Subsystem and Vehicular Communication Gateway ........................ 160 Wei-Kuo Chiang, National Chung Cheng University, Chiaya, Taiwan, R.O.C. An-Nie Ren, National Chung Cheng University, Chiaya, Taiwan, R.O.C. Section 5 Vehicular Ad Hoc Networks and Delay Tolerant Vehicular Networks Chapter 12 MAC Protocols in Vehicular Ad Hoc Networks ................................................................................. 183 Chih-Yung Chang, Tamkang University, Taiwan, R.O.C. Chapter 13 Routing Protocol in Vehicular Ad Hoc Networks ............................................................................... 206 Yuh-Shyan Chen, National Taipei University, Taipei, Taiwan, R.O.C. Yun-Wei Lin, National Taipei University, Taipei, Taiwan, R.O.C. Chapter 14 Applications in Vehicular Ad Hoc Networks ...................................................................................... 229 Tzung-Shi Chen, National University of Tainan, Tainan, Taiwan, R.O.C. Hua-Wen Tsai, National University of Tainan, Tainan, Taiwan, R.O.C. Yi-Shiang Chang, National University of Tainan, Tainan, Taiwan, R.O.C. Chapter 15 DTN Technologies for Vehicular Networks........................................................................................ 252 Kun-Chan Lan, National Cheng Kung University, Tainan, Taiwan, R.O.C. Section 6 Management and Traffic Control Chapter 16 Simple Transporation Management Framework................................................................................. 271 Chyi-Ren Dow, Feng Chia University, Taiwan, R.O.C. Chapter 17 Vehicular System Management Architecture and Application Platform ............................................ 290 Teng-Wen Chang, National Taiwan University of Science and Technology, Taiwan, R.O.C. Jiann-Liang Chen, National Taiwan University of Science and Technology, Taiwan, R.O.C. Chapter 18 Remote Vehicular System Management Functions and Information Structure .................................. 310 Teng-Wen Chang, National Taiwan University of Science and Technology, Taiwan, R.O.C. Jiann-Liang Chen, National Taiwan University of Science and Technology, Taiwan, R.O.C. Chapter 19 Using Wireless Mesh Network for Traffic Control ............................................................................. 331 Kun-Chan Lan, National Cheng Kung University, Tainan, Taiwan, R.O.C. Section 7 Mobility Model, Simulation, and Security Chapter 20 Mobility Models of Vehicular Networks ............................................................................................ 348 Kun-Chan Lan, National Cheng Kung University, Tainan, Taiwan, R.O.C. Chapter 21 MOVE: A Practical Simulator for Mobility Model in VANET .......................................................... 355 Kun-Chan Lan, National Cheng Kung University, Tainan, Taiwan, R.O.C. Chapter 22 Security Attacks of Vehicular Networks ............................................................................................. 369 Jen-Chun Chang, National Taipei University, Taiwan, R.O.C. Chun-I Fan, National Sun Yat-sen University, Taiwan, R.O.C. Ruei-Hau Hsu, National Sun Yat-sen University, Taiwan, R.O.C. Compilation of References ............................................................................................................... 380 About the Contributors .................................................................................................................... 398 Index ................................................................................................................................................... 405 Detailed Table of Contents Foreword ............................................................................................................................................ xix Preface ................................................................................................................................................ xxi Acknowledgment ............................................................................................................................... xxv Section 1 Introduction of Vehicular Networks and Intelligent Transporation Systems Chapter 1 Introduction of Vehicular Network Architectures ................................................................................... 1 Ming-Chiao Chen, National Taitung University, Taitung, Taiwan, R.O.C. Teng-Wen Chang, National Taiwan University, Taipei, Taiwan, R.O.C. A vehicular network organizes and connects vehicles with each other, and with mobile and fixed-locations resources. This chapter discusses the architectures in the vehicular network environment. We introduce the overview of in-vehicle and out-vehicle network architectures. An automobile in an in-vehicle network adopts four vehicle bus protocols, CAN (Controller Area Network), LIN (Local Interconnect Network), MOST (Media Oriented Systems Transport) and FlexRay. However, these protocols cannot intercommunicate with each other. Therefore, the OSEK operating system is designed as standard software architecture for the various ECUs (Electronic Control Units). In the out-vehicle network, the OBU (On Board Unit) in the automobile can communicate with the infrastructure via the Internet. We discuss next-generation vehicular network architecture, the modern in-vehicle networks, on-board computers and the Internet, mobile telecommunications and telematics applications in the ground vehicles, and finally, we introduce future desired features. This chapter discusses the architectures in vehicular network environment. Section 1.1 introduces the overview of in-vehicle and out-vehicle network architectures. Section 1.2 describes in-vehicle network architecture for disaster communication network by combining various automotive bus protocols. Section 1.3 describes the out-vehicle network architecture for disaster communication network by combining various wireless LANs. Section 1.4 discusses next-generation vehicular network architecture, the modern in-vehicle networks, on-board computers and the Internet, mobile telecommunications and telematics applications in the ground vehicles, and introduces future desired features. Chapter 2 Introduction of Vehicular Network Applications .................................................................................. 15 Yao-Chung Chang, National Taitung University, Taiwan, R.O.C. Information and Communication Technology (ICT) is concerned with all the technologies that manage, process, and communicate information. It is also named as telematics, combining two words: telecommunications and informatics, which is widely used in the application of Global Positioning System technology integrated with computers and in the mobile communications technology for automotive navigation systems. Table 2.1 and Table 2.2 respectively list the telemetric applications from user’s point of view and the practical applications of vehicular telematics. Four applications of the vehicular network are discussed in this chapter. Section 2.1 introduces the vehicular network application services. Section 2.2 discusses the vehicular network application management. Section 2.3 provides the platform technologies of vehicular network application. Finally, future vehicular network application and deployments are presented in the Section 2.4. Chapter 3 Introduction to ITS and NTCIP ............................................................................................................ 32 Da-Jie Lin, Feng Chia University, Taiwan, R.O.C. Chyi-Ren Dow, Feng Chia University, Taiwan, R.O.C. Intelligent Transportation Systems (ITS) combines high technology and improvements in information systems, communication, sensors, and relevant mathematical methods with the conventional world of surface transportation infrastructure to increase the capacity of transportation systems and to improve the level of services. There are four major goals of ITS, including safety, environmental protection, efficiency, and economy. NTCIP (NTCIP Standard 9001, 2002; DISA et al., 1997) is a set of communications protocols and data definition standards designed for various needs of ITS services and applications. The key goals of the NTCIP open-standards effort are interoperability and interchangeability. Interoperability refers to the ability for multiple devices to work together as a single system and interchangeability refers to the ability to use multiple brands of a device on the same communications channel. Accompanying the social and economic development, traffic congestion and delay have become major issues in most areas around the world. How to use readily available technologies to increase the capacity of transportation systems and to improve the level of service has become one of major solutions to solve transportation problems that people are facing. This is the motivation of Intelligent Transportation Systems development. NTCIP is a set of communications protocols and data definition standards designed for various needs of ITS services and applications. These standards are intended to handle these needs in the two areas: Center-to-Field (C2F) and Center-to-Center (C2C) communications. Section 2 Embedded System Architecture and Communication Protocols Chapter 4 Vehicular Embedded System Architecture............................................................................................ 58 Chung-Ping Young, National Cheng Kung University, Taiwan, R.O.C. Transportation of humans and objects have been playing an important role in our daily lives since civilization first formed and needed new means of reaching destinations. The invention of efficient transportation greatly reduced the time and labor once required and in addition largely extended the living environment that people can reach. The more time and labor for transportation is saved, the more leisure time people will have. Animal-power or natural resources have been the driving force of transportation for a long time. After the steam engine was invented, the automobile started a new era. The mass production of the Ford model T created the modern automobile industry and made the automobile more affordable. The basic structure of the automobile has not changed much, but evolving technologies has kept improving its functions and performance. The construction of traffic networks and mass production of automobiles have made the automobile the most important land based transportation carrier. The usage of automobiles is usually associated with the growth of economy and industry of a nation, so the population ratio that owns automobiles in a developed country is larger than that in a developing country. When the economy grows, vehicle as a transportation tool becomes more affordable and popular, for instance China or India. When people use automobiles in their daily lives, they demand not only mobility, but also safety, comfort and convenience. These are some design factors that manufacturers have to put into aspect when enhancing functions by introducing and developing new technologies. Chapter 5 Data Communications Inside Vehicular Environments ........................................................................ 74 Cheng-Min Lin, Nan Kai University of Technology, Taiwan, R.O.C. Tzong-Jye Liu, Feng Chia University, Taiwan, R.O.C. ZigBee is based on IEEE 802.15.4 which specifies the physical layer and medium access control (MAC) for low-cost and low-power LR-WPAN. The technology can be applied in intelligent key, A/C operation and steering wheel inside vehicles. There are two types of devices in ZigBee, FFD and RFD. A FFD can communicate with RFDs and other FFDs, while a RFD can only communicate with a FFD. In ZigBee physical layer, it follows IEEE 802.15.4 standard and operates in unlicensed RF worldwide (2.4GHz global, 915MHz Americas or 868 MHz Europe). A superframe contained an active portion and an inactive portion is used in the MAC layer of ZigBee. The active portion includes CAP and CFP. In the inactive partition, the coordinator can enter sleep mode to save its power. Three main topologies of ZigBee are star, mesh, and tree. However, ZigBee is successfully produced into a low-cost controller applied for automotive applications, including vehicle control and status monitoring. According to the forecast of ON World in 2005 (ON WORLD, 2009) , the deployed wireless sensing network nodes will increase to 127 million in 2010 from 1.2 million in 2005. It can be applied in home automation, battlefield surveillance, health care applications and vehicular environments. A wireless sensor network (WSN) constitutes a lot of wireless sensing nodes. In addition, a node in WSN consists of one or more sensors, a radio transceiver, and a microcontroller. The sensor can be used for sensing temperature, pressure, sound, vibration, motion or position, etc. to collect status from devices or environments. The transceiver is used to relay the information of the collected status computed by the microcontroller to a center node, called a gateway or sink. Therefore, a WSN belongs to one type of wireless ad-hoc networks. However, the nodes in a WSN are usually smaller than that in traditional wireless ad-hoc networks regarding node size, computing power, memory size, and transmission rage. In other words, the transmission ability, computing power, and memory size of WSN nodes are limited. Chapter 6 Wireless Access in Vehicular Environments ......................................................................................... 90 Tzong-Jye Liu, Feng Chia University, Taiwan, R.O.C. Ching-Wen Chen, Feng Chia University, Taiwan, R.O.C. The IEEE 1609 standards define communication for wireless access in vehicular environment (WAVE) services, which enable vehicle-to-vehicle, vehicle-to-roadside, as well as vehicle-to-infrastructure communications. The standard consists of four parts, which are briefly described in this chapter. IEEE 1609.1 describes the WAVE resource manager which specifies the wireless access method in a WAVE environment and allows a remote manager application to establish connection with a resource command processor on an on-board unit. IEEE 1609.2 defines several secure message formats to process messages for WAVE system. The standard covers methods for securing WAVE management messages and application messages, which protects messages from attacks such as eavesdropping, spoofing, alteration, replay, and linkable information to unauthorized parties. IEEE 1609.3 defines network services for WAVE systems, whose network services operate at the network and transport layers of the OSI model and support both the IPv6 traffics and the WAVE short message services. IEEE 1609.4 describes WAVE multi-channel operations. It specifies the functions of MAC sublayer management entity and WAVE MAC with channel coordination. The multi-channel operation provides an efficient mechanism that controls the operation of upper layer across multiple channels. Section 3 Location Based Services Chapter 7 Introduction To Global Satellite Positioning System (GPS) ............................................................... 108 Jenq-Muh Hsu, National Chiayi University, Chiayi, Taiwan, R.O.C. Understanding the right positions and directions of people and objects is a significant issue from the ancient eras to the present. In the past, people often launched a war in order to satisfy the craving for the dominating powers and spread their realms. In the recent, Global Satellite Positioning System (GPS) has become the one of most popular positioning technologies. GPS can provide users precise positioning information, no matter wherever that may present their own positions. The early GPS positioning technology has been widely used in military, marine use, until recently gradually applied into our daily life, e.g., automotive navigation, geodesy surveying, etc. In this chapter, the authors will briefly introduce some GPS issues including the origins of GPS, GPS system architecture, and related GPS applications. Chapter 8 Vehicle Location and Navigation Systems ......................................................................................... 119 Ben-Jye Chang, National Yunlin University of Science and Technology, Yunlin, Taiwan, R.O.C. The most driving purpose is to traverse to the destination safely, efficiently, and comfortably. Two types of approaches could achieve the goals, including the static and dynamic approaches. In the static aspect, vehicles use the static road and traffic information to navigate. Conversely, in the dynamic aspect, vehicles adopt the dynamic information instead. However, both of the two approaches first require getting the vehicle’s location and then map the position on an e-map. Thus, this chapter first introduces some important vehicle location determination algorithms: the dead reckoning and global position system algorithms, in which the precision of location technologies are compared. Then, the map-matching algorithm is described in detail. Finally, various vehicle navigation approaches are detailed, in which the important topics include: the navigation architecture, the navigation routing algorithm, and navigation applications. Chapter 9 Design and Implementation of Vehicle Navigation Systems .............................................................. 131 Min-Xiou Chen, National Dong-Hwa University, Hualien, Taiwan, R.O.C. Vehicle Navigation System (VNS) is a complicated and integrated system. A reliable vehicle navigation system should integrate the wireless communication technologies, positioning technologies, embedded computer, geographic information database, and so on. The major purpose of the chapter is to help understanding the architecture of vehicle navigation system. This chapter first introduces the system requirements and system analysis, and show the system platform of vehicle navigation system. The system platform can be divided into six components. There are the digital map database, positioning devices, map-matching process, route planning process, route guidance process, human-machine interface, and wireless communication interface. The design issues and system communication of these components are detail illustrated in the chapter. Finally, the authors also present some vehicle navigation systems proposed in the past few years, and show the difference of these systems. The aim of vehicle navigation system is to guide the vehicle along the optimal path from the starting point to destination. A reliable vehicle navigation system can reduce the traffic chaos in the city and improve the transportation delay. In order to achieve reliable vehicle navigation system, the detail system requirements, system analysis, and system architecture are shown in the chapter. Each component of vehicle navigation system is briefly illustrated, and the system communication is also described. They also present the architecture of the proposed vehicle navigation system, and show the difference of these systems. Therefore this chapter helps understanding the architecture of vehicle navigation system. Section 4 Integrated Vehicular Application Chapter 10 Vehicular Metropolitan Area Network Systems Architecture: The WiMAX Network Reference Model ................................................................................................................................. 144 Cheng Hsuan Cho, National Chung Cheng University, Taiwan, R.O.C. Jen-Yi Pan, National Chung Cheng University, Taiwan, R.O.C. The WiMAX NWG develops a network reference model to serve as an architecture framework for WiMAX deployments and to ensure interoperability among various WiMAX equipment and operators. The network reference model envisions unified network architecture for supporting fixed, nomadic, and mobile deployments and is based on an IP service model. We introduce WiMAX network architecture, WiMAX network entry, mobility management, QoS functional elements, core network planning and accounting architecture in this section. However, all of them are significant in deploying WiMAX core network. The operator tries to reach the goals including system performance, reliability, and so on. On the other hand, the WiMAX operator should consider and balance such many variables in order to achieve a better situation. Chapter 11 Interworking of IP Multimedia Subsystem and Vehicular Communication Gateway ........................ 160 Wei-Kuo Chiang, National Chung Cheng University, Chiaya, Taiwan, R.O.C. An-Nie Ren, National Chung Cheng University, Chiaya, Taiwan, R.O.C. In recent years, more and more people dream of experiencing various IP-based multimedia application services when they are driving through their car. However, those multimedia devices in the car may use different communication protocols such as X.10, Havi, Jini, UPnP and SIP. In order to provide a variety of IP-based multimedia services to those users in the car, we mainly investigate the issue of interworking between IP Multimedia Subsystem (IMS) and telematics of the vehicular industry. A service-integrated platform, Open Service Gateway Initiative Service Platform (OSGi SP), has been proposed to act as a Residential Gateway (RGW) and to administer the communication between the vehicular environment and Internet. Besides, a Home IMS Gateway (HIGA), which can be implemented on a NGN RGW, has been developed by Home Gateway Initiative (HGI) since 2005 to collect the relevant information of in-car users, devices and services and to manage the IMS sessions for the in-car devices that do not support IMS functions. With these techniques, the users can enjoy their digital life by interacting with the home/vehicular network from anywhere. Section 5 Vehicular Ad Hoc Networks and Delay Tolerant Vehicular Networks Chapter 12 MAC Protocols in Vehicular Ad Hoc Networks ................................................................................. 183 Chih-Yung Chang, Tamkang University, Taiwan, R.O.C. With the rapid development of wireless technologies, the Vehicular Ad Hoc Networks (VANETs) have recently received much attention. VANETs technologies aim to ensure traffic safety for drivers, provide comfort for passengers and reduce transportation time and fuel consumption with many potential applications. The achievement of these aims highly relies on efficient MAC protocols which determine the performance of packet transmission in terms of success rate, delay, throughput and bandwidth utilization. This chapter reviews the existing MAC protocols developed for VANETs. Initially, the IEEE 802.11p and DSRC standard are reviewed. Two TDMA-based MAC protocols, called CVIA and VeSOMAC, are then introduced. In addition, three MAC protocols that cope with the emergency-message broadcasting problem are proposed. Finally, a reliable MAC protocol which is developed based on the cluster topology is reviewed. Chapter 13 Routing Protocol in Vehicular Ad Hoc Networks ............................................................................... 206 Yuh-Shyan Chen, National Taipei University, Taipei, Taiwan, R.O.C. Yun-Wei Lin, National Taipei University, Taipei, Taiwan, R.O.C. Vehicular Ad hoc Network (VANET), a subclass of mobile ad hoc networks (MANETs), is a promising approach for the intelligent transportation system (ITS). The design of routing protocols in VANETs is important and necessary issue for support the smart ITS. The key difference of VANET and MANET is the special mobility pattern and rapidly changeable topology. It is not effectively applied the existing routing protocols of MANETs into VANETs. In this chapter, the authors mainly survey new routing results in VANET. They introduce unicast protocol, multicast protocol, geocast protocol, mobicast protocol, and broadcast protocol. It is observed that carry-and-forward is the new and key consideration for designing all routing protocols in VANETs. With the consideration of multi-hop forwarding and carryand-forward techniques, min-delay and delay-bounded routing protocols for VANETs are discussed in VANETs. Besides, the temporary network fragmentation problem and the broadcast storm problem are further considered for designing routing protocols in VANETs. The temporary network fragmentation problem caused by rapidly changeable topology influence on the performance of data transmissions. The broadcast storm problem seriously affects the successful rate of message delivery in VANETs. The key challenge is to overcome these problems to provide routing protocols with the low communication delay, the low communication overhead, and the low time complexity. Chapter 14 Applications in Vehicular Ad Hoc Networks ...................................................................................... 229 Tzung-Shi Chen, National University of Tainan, Tainan, Taiwan, R.O.C. Hua-Wen Tsai, National University of Tainan, Tainan, Taiwan, R.O.C. Yi-Shiang Chang, National University of Tainan, Tainan, Taiwan, R.O.C. The various sensors and wireless communication devices have been extensively applied to daily life due to the advancements of microelectronics mechanism and wireless technologies. Recently, vehicular communication systems and applications become more and more important to people in daily life. Vehicular communication systems that can transmit and receive information to and from individual vehicles have the potential to significantly increase the safety of vehicular transportation, improve traffic flow on congested roads, and decrease the number of people of deaths and injuries in vehicular collisions effectively. This system relies on direct communication between vehicles to satisfy the communication needs of a large class of applications, such as collision avoidance, passing assistance, platooning. In addition, vehicular communication systems can be supplemented by roadside infrastructure to access Internet and other applications. This system forms a special case of mobile ad hoc networks called Vehicle Ad Hoc Networks (VANETs). They can be formed between vehicles with vehicle to vehicle (V2V) communication or between vehicles and an infrastructure with vehicle to infrastructure (V2I) communication. The applications and characteristics of VANETs are introduced and presented in this Chapter. Chapter 15 DTN Technologies for Vehicular Networks........................................................................................ 252 Kun-Chan Lan, National Cheng Kung University, Tainan, Taiwan, R.O.C. A Delay Tolerant Network (DTN) is one type of challenged network where network contacts are intermittent or link performance is highly variable or extreme. In such a network, a complete path does not exist from source to destination for most of the time. In addition, the path can be highly unstable and may change or break unexpectedly. To make communication possible in a delay tolerant network, the intermediate nodes need to take custody of data during the blackout and forward it toward the destination when the connectivity resumes. A vehicular network nicely falls into the context of DTN since the mobility of vehicles constantly causes the disruption of link connectivity’s between vehicles. In this chapter, the authors discuss some research challenges and issues which might occur in a Delay Tolerant Network and how they are related to vehicular networks. Section 6 Management and Traffic Control Chapter 16 Simple Transporation Management Framework................................................................................. 271 Chyi-Ren Dow, Feng Chia University, Taiwan, R.O.C. The Simple Transportation Management Framework (STMF) specifies a set of rules and protocols which can be used to organize, describe, and exchange transportation management information between transportation management applications and equipments. The STMF framework consists of four elements, including Management Information Base (MIB), Structure and Identification of Management Information (SMI), Simple Network Management Protocol (SNMP), and Simple Transportation Management Protocol (STMP). MIB is a collection of management objects written in ASN.1 notation. SMI is the definition of how to create management objects and a hierarchical definition of nodes where management objects will be attached for unique identification. SNMP is a communications protocol for configuring and monitoring of network devices. STMP is a variation of SNMP to address low-bandwidth communication links and real-time device monitoring. Chapter 17 Vehicular System Management Architecture and Application Platform ............................................ 290 Teng-Wen Chang, National Taiwan University of Science and Technology, Taiwan, R.O.C. Jiann-Liang Chen, National Taiwan University of Science and Technology, Taiwan, R.O.C. Notably, not all telematics services can be used in telematics terminals as a result of the varied platform standards. The main issues are that most telematics technologies depend on vertical, proprietary and closed per-OEM Original Equipment Manufacture (OEM) platforms, forming islands of non-interoperable technology and preventing third-party service providers from creating valuable services. In this study, the Open Gateway Service Initiative Vehicle Expert Group (OSGi/VEG) was integrated into an Android platform to generate a vehicular Android/OSGi platform that has the advantages of both original platforms, such as remote management, rich class sharing, proprietary vehicular applications, security policies, easy management of application programming interface (APIs), and an environment with increased openness. Furthermore, this study integrates the cloud computing mechanism into the Android/OSGi platform, which allows service providers to upload their telematics bundles onto storage clouds via the provisioning server. Chapter 18 Remote Vehicular System Management Functions and Information Structure .................................. 310 Teng-Wen Chang, National Taiwan University of Science and Technology, Taiwan, R.O.C. Jiann-Liang Chen, National Taiwan University of Science and Technology, Taiwan, R.O.C. Due to the rapid development of information technology, the network has already spread to every corner of vehicle. With all kinds of ECU devices appear in the vehicle, and it brings the more and more convenient living. On purpose solving heterogamous technologies that are incompatible with each other, developed a “WBEM-based Remote Management and Heterogeneous Vehicular Network Diagnosis System” on OSGi Gateway. This system can focus on a variety of problems come from vehicle network, and find out what are the problems or where are the problems happened. If the problem still can not be solved properly, we must to seek for help from remote managers. The users can acquire enough information without understanding how to control every device, so that the users can help near diagnosis system to solve vehicle network’s problems and to promote the abilities of near network diagnosis. Chapter 19 Using Wireless Mesh Network for Traffic Control ............................................................................. 331 Kun-Chan Lan, National Cheng Kung University, Tainan, Taiwan, R.O.C. Wireless mesh networks (WMN) have attracted considerable interest in recent years as a convenient, flexible and low-cost alternative to wired communication infrastructures in many contexts. However, the great majority of research on metropolitan-scale WMN has been centered around maximization of available bandwidth, suitable for non-real-time applications such as Internet access for the general public. On the other hand, the suitability of WMN for missioncritical infrastructure applications remains by and large unknown, as protocols typically employed in WMN are, for the most part, not designed for real-time communications. In this chapter, the authors describe a real-world testbed, which sets a goal of designing a wireless mesh network architecture to solve the communication needs of the traffic con- trol system in Sydney, Australia. This system, known as SCATS (Sydney Coordinated Adaptive Traffic System) and used in over 100 cities around the world, connects a hierarchy of several thousand devices -- from individual traffic light controllers to regional computers and the central Traffic Management Centre (TMC) - and places stringent requirements on the reliability and latency of the data exchanges. The authors discuss some issues in the deployment of this testbed consisting of 7 mesh nodes placed at intersections with traffic lights, and show some results from the testbed measurements. Section 7 Mobility Model, Simulation, and Security Chapter 20 Mobility Models of Vehicular Networks ............................................................................................ 348 Kun-Chan Lan, National Cheng Kung University, Tainan, Taiwan, R.O.C. A key component for VANET simulations is a realistic vehicular mobility model that ensures conclusions drawn from simulation experiments will carry through to real deployments. However, VANET simulations raise many new questions about suitable levels of details in simulation models. To get accurate results, the mobility models of Vehicular Networks should be as realistic as possible, and involve road-maps with all constraints and facilities related to the vehicular movement. In this chapter, the authors provide an overview of some mobility models that are relevant to VANETs. The criteria of applicability they consider here is the employment of road maps, and thus limiting the nodes movements into the routes, instead of moving them in a wide open area. They compare different models based on the parameters they use. For instance, some models use traffic control mechanisms (stop signs or traffic lights) at route intersections, and some just assume continuous movement at these points. Some assume routes to be single-lane, some others support multi-lanes routes. Some define the security distance, while others just ignore this parameter. Chapter 21 MOVE: A Practical Simulator for Mobility Model in VANET .......................................................... 355 Kun-Chan Lan, National Cheng Kung University, Tainan, Taiwan, R.O.C. Vehicular Ad-Hoc Network (VANET) is surging in popularity, in which vehicles constitute the mobile nodes in the network. Due to the prohibitive cost of deploying and implementing such a system in real world, most research in VANET relies on simulations for evaluation. A key component for VANET simulations is a realistic vehicular mobility model that ensures conclusions drawn from simulation experiments will carry through to real deployments. However, VANET simulations raise many new questions about suitable levels of details in simulation models for nodes mobility. In VANET simulations, the mobility models used affect strongly the simulation output. The researchers need to decide what level of details are required for their simulations. In this chapter, the authors introduce a tool MOVE that allows users to rapidly generate realistic mobility models for VANET simulations. MOVE is built on top of an open source micro-traffic simulator SUMO. The output of MOVE is a realistic mobility model and can be immediately used by popular network simulators such as ns-2 and Qualnet. They show that the simulation results obtained when using a realistic mobility model such as MOVE are significantly different from results based on the commonly used random waypoint model. In addition, they evaluate the effects of details of mobility models in three case studies of VANET simulations (specifically, the existence of traffic lights, driver route choice and car overtaking behavior) and show that selecting sufficient level of details in the simulation is critical for VANET protocol design. Chapter 22 Security Attacks of Vehicular Networks ............................................................................................. 369 Jen-Chun Chang, National Taipei University, Taiwan, R.O.C. Chun-I Fan, National Sun Yat-sen University, Taiwan, R.O.C. Ruei-Hau Hsu, National Sun Yat-sen University, Taiwan, R.O.C. The application of vehicular ad hoc network (VANET) improves driving safety and traffic management. Due to the above applications, security attacks on VANET can be serious threats all the time. VANET is a special form of mobile ad hoc network (MANET). Hence any attacks exist on MANET also can be arisen on VANET. Moreover, some special attacks can be raised on VANET, which don’t exist on MANET. Nevertheless, some characteristics of VANET can be positive effects and some can be negative effects on security issues. Before designing the security mechanism to defend attacks, the authors should take the positive effects and avoid the negative effects on the security of VANET. Furthermore, they class all possible attacks of VANET from every network layer. The authors also introduce the reason of forming every attack and the possible effect on VANET in detail. Therefore this chapter helps understanding the latent threats and the useful resources of security issues on VANET. Compilation of References ............................................................................................................... 380 About the Contributors .................................................................................................................... 398 Index ................................................................................................................................................... 405 xix Foreword The Human Resource Program for Information and Communication Technology sponsored by the Ministry of Education (MOE) in Taiwan, lead by myself, has been a key program in the past 10 years to bridge possible gaps between university education and industrial human resource demands, and to train and cultivate sufficient as well as high-quality and skilled young professionals in the blooming global telecommunication market. After a detailed analysis of the ICT industrial trend by a review board consisted of experts and professionals from both academic and industrial sector around 2006, MOE in Taiwan decided to focuses the talent cultivation direction only on a few selected areas of emerging technology. Telematics and Vehicular Networks has emerged as one such important technology since the research progress in the telematics and vehicular networking was so significant that related industry has become booming around the world and one can easily expect future drivers can enjoy the benefit from related smart telematics products and services. We also believe telematics and vehicular networks can be categorized as a special kind of Green ICT Technologies since it also helps to save energy in many scenarios. In the past few decades, Taiwan’s telecommunication and ICT industry has experienced a long period of high growth and fast technology evolution. For example, the communication industry in Taiwan has increased up to 8 folds around 10 years, with its manufacturing capacity ranging from traditional LAN switches to 3G smart phones. Without surprises, recent industry trends in smart cars and vehicular networks have also created a strong demand on talent engineers with good hand-on experiences in related products and services. Again, universities and the academic community have been asked to keep upgrading key ICT courses and laboratories to link up with the telematics industry in a timely fashion. In response to this demand, a prestigious team was selected from the academic community in Taiwan in 2007, to aggregate teaching resources, refine the essential courseware, and enhance experiment environments for training talented students in this field. An intercollegiate telematics promotion center was also established for completing this task and Professor Chung-Ming Huang, National Cheng-Kung University, was selected to lead this center for his dedication and the knowledge and research experience he has accumulated in this field. This book, Telematics Communication Technologies and Vehicular Networks: Wireless Architectures and Applications, edited by Prof. Chung-Ming Huang and Prof. Yuh-Shyan Chen, is a work contributed by such a group of telematics experts and professors in this field. This book has successfully covered a wide range of technical topics, including vehicular network architecture, related communication protocols, ITS/telematics applications, navigation systems, location based services and embedded systems. xx Many chapters in this book is self-guided and can be used a tutorial. In general, it should be a valuable textbook, guidance and/or reference for students, researchers, engineers, technical managers, and other professionals in this field. I believe all readers can enjoy reading this book. Zsehong Tsai, Professor Human Resource Program Office for ICT National Taiwan University, Taiwan, R.O.C. xxi Preface TelemaTics communicaTion Technologies and Vehicular neTworks: wireless archiTecTures and applicaTions Telematics communication technologies and vehicular networks have been identified as key technologies for increasing road safety and transport efficiency. Telematics communication technologies and vehicular networks aim to ensure traffic safety for drivers, provide comfort for passengers and reduce transportation time and fuel consumption. The development of vehicular communication and networking technologies are expected to enable many potential applications, including automatic collision notification and prevention, emergency management, assistances for safe driving, real-time traffic congestion notification, location-based driver information services, high-speed tolling, vehicle tracking, automobile Internet access, and many others. To facilitate these applications, many different new types of communication and networking would be involved, including intra-vehicle, vehicle-to-vehicle, vehicle-to-roadside and vehicle-to-infrastructure communications. This book aims to provide a fast and complete view of all aspects related to telematics communication technologies and vehicular networks. This book is intended for graduate students, researchers, engineer, and practitioners who are interested in acquiring a global view of the set of techniques and protocols that have been referred to as “Telematics communication technologies and vehicular networks: wireless architectures and applications” in the literature. The book can serve as a reference resource for researchers, engineers, and developers working in the field of telematics technologies. This book includes 22 chapters which are classified into 7 Sections. Section 1 introduces the vehicular networks and intelligent transportation systems. Section 2 describes embedded system architecture and communication protocols. Section 3 reports location based services. Section 4 provides integrated vehicular applications. Section 5 presents vehicular ad hoc networks and delay tolerant vehicular networks. Section 6 explains management and traffic control. Finally, Section 7 talks about mobility model, simulation, and security for telematics communication technologies and vehicular networks. The first section of the book, Introduction of Vehicular Networks and Intelligent Transportation Systems, presents introductory materials that is preparatory for what us described in the rest of the book. Chapter 1, by Ming-Chiao Chen and Teng-Wen Chang, gives a short introduction to vehicular network architectures. This chapter discusses the architectures in the vehicular network environment, and introduces the overview of in-vehicle and out-vehicle network architectures. This chapter also discusses the next-generation vehicular network architecture, the modern in-vehicle networks, on-board computers and the Internet, mobile telecommunications and telematics applications in the ground vehicles. Chapter 2, by Yao-Chung Chang, introduces vehicular network applications. Four applications of the vehicular network are surveyed and discussed in this chapter. They are vehicular network application xxii services, vehicular network application managements, the platform technologies of vehicular network application, and the future vehicular network application and deployment. Chapter 3, by Da-Jie Lin and Chyi-Ren Dow, explains intelligent transportation systems. Intelligent Transportation Systems (ITS) combines high technology and improvements in information systems, communication, sensors, and relevant mathematical methods with the conventional world of surface transportation infrastructure to increase the capacity of transportation systems and to improve the level of services. There are four major goals of ITS, including safety, environmental protection, efficiency, and economy. The second section of the book, Embedded System Architecture and Communication Protocols, presents the vehicular embedded system architecture, data communication protocols for vehicular network, and wireless access techniques in vehicular environments. Chapter 4, by Chung-Ping Young, explains the vehicular embedded system architecture. The car electronics plays an increasingly important role in automobile industry. The embedded system has already been extensively employed for improving the operation and performance of vehicles, such as safety, comfort, convenience, and environmental protection. In terms of electronic system, an automobile is a distributed embedded system, and the control messages to each electronic control unit (ECU), go through the in-vehicle network. An ECU is an embedded processor or computing system, integrated with a data acquisition device or an electromechanical driver. Chapter 5, Cheng-Min Lin, and Tzong-Jye Liu, reports the data communications inside vehicular environments. ZigBee is based on IEEE 802.15.4 which specifies the physical layer and medium access control (MAC) for low-cost and low-power LR-WPAN. ZigBee is successfully produced into a low-cost controller applied for automotive applications, including vehicle control and status monitoring. Chapter 6, by Tzong-Jye Liu and Ching-Wen Chen, gives the wireless access in vehicular environments. This chapter describes the IEEE 1609 standard for Wireless Access in Vehicular Environment (WAVE) services, which enable vehicle-to-vehicle, vehicle-to-roadside, as well as vehicle-to-infrastructure communications. The standard consists of four parts, IEEE 1609.1, IEEE 1609.2, IEEE 1609.3, and IEEE 1609.4, which are described in this chapter. The third section of the book, Location Based Services, provides the useful knowledge and technique of location based services for supporting telematics communication technologies and vehicular networks. Chapter 7, by Jenq-Muh Hsu, provides the introduction of GPS. Global Satellite Positioning System (GPS) recently has become the one of most popular positioning technologies. This chapter briefly introduces some GPS issues including the origins of GPS, GPS system architecture, and related GPS applications. Chapter 8, by Ben-Jye Chang, presents the vehicle location and navigation systems. This chapter introduces important vehicle location determination algorithms: the dead reckoning and global position system algorithms, in which the precision of location technologies are compared. The map-matching algorithm is then described. Various vehicle navigation approaches are explained. Chapter 9, by Min-Xiou Chen, discusses the design and implementation of vehicle navigation systems. The major purpose of the chapter is to understand the architecture of vehicle navigation system. This chapter introduces the system requirements and system analysis, and show the system platform of vehicle navigation system. In the fourth section of the book, Integrated Vehicular Application, presents the vehicular metropolitan area network systems architecture and the interworking of IP multimedia subsystem and vehicular communication gateway. xxiii Chapter 10, by Cheng Hsuan Cho and Jen-Yi Pan, reports the vehicular metropolitan area network systems architecture: the WiMAX network reference model. This chapter introduces WiMAX network architecture, WiMAX network entry, mobility management, QoS functional elements, core network planning and accounting architecture. WiMAX technique is the one of important wireless access techniques for the vehicular communication. Chapter 11, by Wei-Kuo Chiang and An-Nie Ren, presents the interworking of IP multimedia subsystem and vehicular communication gateway. To provide a variety of IP-based multimedia services to those users in the car, this chapter investigates the issue of interworking between IP Multimedia Subsystem (IMS) and telematics of the vehicular industry. Section 5 of the book, Vehicular Ad Hoc Networks and Delay Tolerant Vehicular Networks, presents MAC, routing protocols, and applications for vehicular ad hoc networks, and DTN technologies for vehicular networks. Chapter 12, by Chih-Yung Chang, presents MAC protocols in vehicular ad hoc networks. The achievement of requirements of VANETs highly relies on efficient MAC protocols which determine the performance of packet transmission in terms of success rate, delay, throughput and bandwidth utilization. This chapter reviews existing MAC protocols developed for VANETs. Chapter 13, by Yuh-Shyan Chen and Yun-Wei Lin, presents routing protocols in vehicular ad hoc networks. The design of routing protocols in VANETs is important and necessary issue for support the smart ITS. This chapter discusses existing routing results, including unicast protocol, multicast protocol, geocast protocol, mobicast protocol, and broadcast protocol, in VANETs. Chapter 14, by Tzung-Shi Chen, Hua-Wen Tsai, and Yi-Shiang Chang, discusses applications in vehicular ad hoc networks. Vehicular communication systems and applications become more and more important to people in daily life. The applications of VANETs are introduced in this Chapter. Chapter 15, by Kun-Chan Lan, reports DTN technologies for vehicular networks. A Delay Tolerant Network (DTN) is one type of challenged network where network contacts are intermittent or link performance is highly variable or extreme. This chapter discusses some research challenges and issues which might occur in a Delay Tolerant Network. Section 6 of the book, Management and Traffic Control, presents the simple transportation management framework, the vehicular system management architecture and application platform, the remote vehicular system management function and information structure, and the wireless mesh network for the traffic control. Chapter 16, by Chyi-Ren Dow, discusses the simple transportation management framework. The Simple Transportation Management Framework (STMF) specifies a set of rules and protocols which can be used to organize, describe, and exchange transportation management information between transportation management applications and equipments. Chapter 17, by Teng-Wen Chang and Jiann-Liang Chen, reports the vehicular system management architecture and application platform. In this chapter, the Open Gateway Service Initiative Vehicle Expert Group (OSGi/VEG) was integrated into an Android platform to generate a vehicular Android/ OSGi platform. Chapter 18, by Teng-Wen Chang and Jiann-Liang Chen, presents the remote vehicular system management functions and information structure. On purpose solving heterogamous technologies that are incompatible with each other, this chapter develops a “WBEM-based Remote Management and Heterogeneous Vehicular Network Diagnosis System” on the OSGi Gateway. Chapter 19, by Kun-Chan Lan, presents the wireless mesh network for traffic control. Wireless mesh networks (WMN) have attracted considerable interest in recent years. This chapter described a real-world xxiv testbed, which sets a goal of designing a wireless mesh network architecture to solve the communication needs of the traffic control system in Sydney, Australia. The final section of the book, Mobility Model, Simulation, and Security, provides a detailed description of mobility models of vehicular networks, REALISTIC simulation of vehicular networks, and security attacks of vehicular networks. Chapter 20, by Kun-Chan Lan, reports the mobility models of vehicular networks. A key component for vehicular network simulation is a realistic vehicular mobility model that ensures conclusions drawn from simulation experiments will carry through to real deployments. To get accurate results, the mobility models of vehicular networks should be as realistic as possible, and involve road-maps with all constraints and facilities related to the vehicular movement. Therefore, this chapter provides an overview of some mobility models that are relevant to vehicular networks. Chapter 21, by Kun-Chan Lan, presents the realistic simulation of vehicular networks. This chapter introduces a tool, MOVE, that allows users to rapidly generate realistic mobility models for vehicular network simulations. MOVE is built on top of an open source micro-traffic simulator SUMO. The output of MOVE is a realistic mobility model and can be immediately used by popular network simulators such as ns-2 and Qualnet. Chapter 22, by Jen-Chun Chang, Chun-I Fan, and Ruei-Hau Hsu, discusses the security attacks of vehicular networks. This chapter classifies all possible attacks of vehicular network from every network layer, and also introduces the reason of forming every attack and the possible effect on vehicular networks. Finally, we thank all contributors of the book for their outstanding contributions. We hope you will enjoy reading this book as we did and you will find this issue informative and helpful in keeping yourselves up-to-date in the fast changing field of telematics communication technologies and vehicular networks, from wireless architectures to applications Chung-Ming Huang National Cheng Kung University, Taiwan, R.O.C. Yuh-Shyan Chen National Taipei University, Taiwan, R.O.C. July 2009 xxv Acknowledgment The editors and authors are working members of The Promotion Center for Telematics Consortium (PCTC), which is part of the Information and Communication Human Resource Program, Ministry of Education (MOE), Taiwan, R.O.C. Thanks for the program of MOE such that the editors and authors can be grouped together. Many thanks to all authors for their hard work and cooperation for delivering their chapters. We also would like to thank Joel A. Gamon of IGI Global for his help and encouragement during this period. Section 1 Introduction of Vehicular Networks and Intelligent Transporation Systems 1 Chapter 1 Introduction of Vehicular Network Architectures Ming-Chiao Chen National Taitung University, Taitung, Taiwan, R.O.C. Teng-Wen Chang National Taiwan University, Taipei, Taiwan, R.O.C. absTracT A vehicular network organizes and connects vehicles with each other, and with mobile and fixed-locations resources. This chapter discusses the architectures in the vehicular network environment. The authors introduce the overview of in-vehicle and out-vehicle network architectures. An automobile in an in-vehicle network adopts four vehicle bus protocols, CAN (Controller Area Network), LIN (Local Interconnect Network), MOST (Media Oriented Systems Transport) and FlexRay. However, these protocols cannot intercommunicate with each other. Therefore, the OSEK operating system is designed as standard software architecture for the various ECUs (Electronic Control Units). In the out-vehicle network, the OBU (On Board Unit) in the automobile can communicate with the infrastructure via the Internet. The authors discuss next-generation vehicular network architecture, the modern in-vehicle networks, on-board computers and the Internet, mobile telecommunications and telematics applications in the ground vehicles, and finally, we introduce future desired features. This chapter discusses the architectures in vehicular network environment. The first section introduces the overview of in-vehicle and out-vehicle network architectures. The next section describes in-vehicle network architecture for disaster communication network by combining various automotive bus protocols. The third section describes the out-vehicle network architecture for disaster communication network by combining various wireless LANs. The last section discusses next-generation vehicular network architecture, the modern in-vehicle networks, on-board computers and the Internet, mobile telecommunications and telematics applications in the ground vehicles, and introduces future desired features. DOI: 10.4018/978-1-60566-840-6.ch001 Copyright © 2010, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global is prohibited. Introduction of Vehicular Network Architectures inTroducTion A vehicular network organizes and connects vehicles with each other, and with mobile and fixed-locations resources (Wu et al., 2005). Many telematics architectures, including navigation services architecture, traffic information architecture, location-based services architecture, entertainment services architecture, emergency and safety services architecture have been provided. In these architectures, traffic information and navigation services are generally provided by central TSPs (Telematics Service Providers). Emergency and safety services are supplied by an on-board platform, which is likely to be installed by the individual car manufacturers. In contrast to these conventionally adopted architectures, telematics architectures are rarely applied in public local hotspots such as public parking lots, hotels, restaurants, airports and shopping centers. In local hot-spot architecture, a vehicle is considered as an alternative mobile computing platform (logically equivalent to a PDA, laptop or a cellular phone) with short-range localized WLAN (Wireless LAN) devices such as Bluetooth and Wi-Fi. This local (hot-spot) architecture allows the car driver to interact with many local services. Telematics architectures will be useful for telematics services only for vehicles providing traditional services such as traffic information, navigation services provided by the central TSP, and local services provided by distributed third-party service providers that can supply the appropriate contextual data. For instance, consider a driver wishing to drive a car to a convention center. The driver initially finds routes to the center using a navigator, then selects a route free from traffic jam based on traffic information from a TSP. The car automatically discovers the resources of the convention center, obtains directions to the designated parking lot, and makes associated payments using the WLAN communication, as it enters the premises. The driver can obtain various local services provided 2 by the convention center even before stepping out of the car. Current telematics systems depend on mobile infrastructures to deliver telematics services to service users. Therefore, deploying a telematics service between mobile networks it is a very expensive task. System developers need to have strong knowledge about the underlying mobile network. Additionally, a telematics terminal cannot be applied for the telematics service from another telematics service provider, since telematics service developers devise their own protocols between telematics terminals and a service provider. Figure 1 shows an overview of in-vehicle network architecture and out-vehicle network architecture. An automobile in an in-vehicle network adopts four vehicle bus protocols, CAN (Controller Area Network), LIN (Local Interconnect Network), MOST (Media Oriented Systems Transport) and FlexRay. However, these protocols cannot intercommunicate with each other. Therefore, the OSEK operating system was designed as standard software architecture for the various ECUs (Electronic Control Units). In the outvehicle network, the OBU (On Board Unit) in the automobile can communicate with the infrastructure via the Internet. The remote home service and remote vehicular service providers provide particular services to an automotive user. The invehicle and out-vehicle network architectures are discussed in detail in the next two sections. in-Vehicle neTwork archiTecTure This section introduces an in-vehicle network architecture for disaster communication network that combine different automotive bus protocols, namely Controller Area Network (CAN), Local Interconnect Network (LIN) and the recently developed FlexRay protocol standard. Moreover, the OSEK/VDX operating system, a joint project Introduction of Vehicular Network Architectures Figure 1. In-vehicle and out-vehicle network architecture of the automotive industry, manipulates automotive messages between different bus protocols to support efficient usage of resources for automotive control unit application software. controller area network German automotive system supplier Robert Bosch created CAN in the mid-1980s for automotive applications as an effective means of allowing robust serial communication (Pazul et al., 1999). The goal was to establish a standard for more reliable and efficient communication by integrating devices, sensors and actuators in a system for real-time control applications. The CAN protocol combines networks and electronic control units thus reducing both wiring harness and complexity. CAN has now gained widespread use in automotives and mobile applications, as well as in industrial automation applications. The CAN (Robert Bosch et al., 1991) specification has two parts. Part A describes the CAN message format as it is defined in CAN specification 1.2, and part B describes both standard and extended message formats. To achieve design transparency and implementation flexibility, CAN is divided into three layers, the object, transfer and physical layers. Figure 2 Johansson et al. (2003) illustrates a CAN bus with three nodes. Johansson et al. (2003) described an application process example of a node. A temperature sensor decides when to request the transmission of a message frame. The frame consists of a data field and overhead, such as the identifier and control fields. Since the application processes are asynchronous, the bus has a mechanism called CSMA/CD, carrier sense multiple access/collision avoidance, for resolving conflicts. The protocol listens to the network in order to avoid collision. CAN has four frame types, namely data, remote, error and overload frames. The data frame is the only frame that is adopted for data transmission. The data frame has two message formats, base frame format (with 11-bit identifier) and 3 Introduction of Vehicular Network Architectures Figure 2. Three nodes connected through a CAN bus sub-network of a CAN bus to integrate intelligent sensor devices or actuators in modern cars. The main features of LIN include: • • • extended frame format (with 29-bit identifier). This discussion focuses on the base frame format, shown in Figure 3 Johansson et al. (2003). The start-of-frame (SOF) bit denotes the start of the frame transmission. It is followed by the 11-bit identifier and the remote transmission request (RTR) bit. The control field consists of 6 bits, and denotes the number of bytes of data that follow in the data field. The data field contains 0-8 bytes. It is followed by the cyclic redundancy checksum (CRC) field, which enables the receiver to check whether the received bit sequence is corrupt. The transmitter adopts 2-bit acknowledgment (ACK) field, ACK slot bit, to receive an acknowledgment of a valid frame from any receiver. The end of a message frame is a 7-bit end-of-frame (EOF). local interconnect network The Local Interconnect Network Bus (LIN-Bus) is a vehicle bus standard or computer networking bus-system used within current automotive network architectures. The LIN specification is enforced by the LIN-consortium, with the first exploited version 1.1, released in 1999. The specification has since evolved to version 2.1 to satisfy current networking needs. The LIN bus is a small and slow network system that is used as a cheap Figure 3.CAN message frame 4 • • • • • • Mono-master, up to 15 slaves 1 wired bus Bitrates 1-20 Kbits/s: 2.4Kbits, 9.6Kbits and 19.2Kbits are usually used in automotive applications Multicast (broadcast) messages Self-synchronization of the slave (only the master has an accurate clock as crystal) Messages with 2,4 or 8 data bytes, and 3 control bytes Error detection by 8 bits checksum and 2 parity bits in identifier Physical layer: ISO9141 Sleep / wake-up capability The LIN implements serial communication in a state-machine, with small microcontrollers or CPLDs. A slave ECU does not need an accurate clock, so can replace crystals or resonators with an RC cell. This is a cost-effective means of designing smart actuators or sensors, or smart connectors. The specification describes three of the seven layers of the OSI model namely physical layer, data link layer and application layer. A LIN network consists of a LIN master and one or more LIN slaves. The LIN bus in an automotive application is generally connected between smart sensors or actuators and an Electronic Control Unit (ECU), which is often a gateway with a CAN bus. A LIN network may have several LIN buses with no interconnection between them, as shown in Figure 4. This network markedly dif- Introduction of Vehicular Network Architectures Figure 4. LIN bus fers from other low-cost buses as K-line which was intended to link all the ECUs to an external analysis approach for diagnosis purposes. Flexray FlexRay is a new communication standard that provides a high-speed serial communication, timetriggered bus and fault-tolerant communication between electronic devices for future automotive applications (FlexRay, 2005; Xu et al., 2008). FlexRay (2005) was developed for the next generation of automobiles and future applications, including x-by-wire, by a consortium founded by BMW, Bosch, DaimlerChrysler and Philips in 2000 (FlexRay Consortium, 2009). The FlexRay protocol provides a high-speed, deterministic and fault-tolerant communications technology. FlexRay is designed specifically for in-vehicle networking, and therefore does not replace existing networks, but instead works in conjunction with already well-established systems, such as the controller area network (CAN), local interconnect network (LIN) and media oriented systems transport (MOST). In Figure 5, an in-vehicle network with FlexRay serving as the backbone provides determinism for engine control and fault tolerance for steer-by-wire, brake-bywire and other advanced safety applications. vasos and osek/VdX in-Vehicle management system vASOS (Sun et al., 2006) (Vehicular Application Specific Embedded Operating Systems) is designed specifically for vehicle use, and is designed to run on a high-performance user interface computer. It fulfills and provides specific Figure 5. FlexRay backbone 5 Introduction of Vehicular Network Architectures device drivers, such as CAN/LIN buses, which are used to communicate with other ECU nodes for diagnostic purposes, and other fundamental network functions for the vehicle domain. OSEK/VDX (Kim et al., 2007) (Open systems and corresponding interfaces for automotive electronics / Vehicle Distributed eXecutive) is an open vehicular industry standard that was founded as a French-German joint project, and which is now drawing attention world-wide. The primary aims of OSEK/VDX are to address the high cost in developing and redeveloping Electronic Control Unit (ECU) software, and to improve the compatibility of those applications. The most important advantages of OSEK/VDX include portability and reliability. OSEK/VDX includes specifications for OSEK Operating System (OS), OSEK Communication (COM), OSEKNetwork Manager (NM) and OSEK Implementation Language (OIL). The properties of vASOS are as follows: • • Focusing on system real-time performance, scalability and robustness. Small kernel size, small memory footprint, low-cost and high efficiency. • • • Plug and play device driver interfaces for expansibility. Emphasis on network control methods, especially wireless networks. Fast boot-up and good power management. Figure 6 shows the OSEK/VDX In-vehicle Management System, which consists of three components: • • vASOS COM: Although OSEK/VDX COM provides a rich set of communication facilities, many applications, such as this vASOS prototype model, probably only require a minimum subset of this functionality and all observe OSEK/VDX COM specification. OSEK/VDX COM module is composed of: ◦ An Interaction layer: The layer that provides communication services for the transfer of application messages. ◦ A Network layer: The layer that provides services for the unacknowledged and segmented transfer of Figure 6. OSEK/VDX in-vehicle network management architecture 6 Introduction of Vehicular Network Architectures • application messages. The network layer provides flow control mechanisms to allow interfacing of communication peers featuring different level of performance and capabilities. ◦ A Data link layer interface that provides services for the unacknowledged transfer of individual data packets over a network to the layers above it. vASOS NM: The NM module focuses mainly on ensuring the safety and the reliability of a communication network for ECUs. This module provides two alternative mechanisms for network monitoring: indirect monitoring by monitored application messages and direct monitoring by dedicated NM communication using the token principle. ouT-Vehicle neTwork archiTecTure This section introduces out-vehicular network architecture for disaster communication network by combining different wireless networks. Currently available wireless technologies, including IEEE802.11p (2007), IEEE802.16 (WiMAX) (2006), and cellular networks, are combined with a mobile router, and loaded on a car to construct a mobile network node. ieee 802.11p A vehicular environment requires a set of new requirements on modern wireless communication systems. Vehicular safety communications applications cannot tolerate long connection establishment delays before communicating with other vehicles encountered on the road. The IEEE 802.11p standard, also referred to as Wireless Access for the Vehicular Environment (WAVE), is designed to solve these issues. The WAVE protocol provides improvements to the physical (PHY) and medium access control (MAC) layers of the existing 802.11 wireless standards. The Department of Transportation of the United States advocates and supports ITS (Intelligent Transportation System) based on dedicated Figure 7. Vehicle safety communication examples 7 Introduction of Vehicular Network Architectures short range communications (DSRC (Xiang et al., 2006)) systems that provide vehicle-to-vehicle and vehicle-to-roadside information exchanges. ITS largely focuses on enabling public safety applications that can save lives and improve traffic flow. Jiang et al. (2006) Two such application scenarios are shown in Figure 7. Private services are also allowed in order to spread the deployment costs, and to encourage the quick development and adoption of DSRC technologies and applications. The DSRC spectrum, as shown in Figure 8 (Eichler et al., 2007; Jiang et al., 2006), is structured into seven channels, each with a bandwidth of 10-MHz. The control channel (CCH) is restricted to safety communications only. The two channels at the ends of the spectrum band are reserved for special applications. The service channels (SCH) is available for both safety and non-safety usage. wimaX The Mobile WiMAX network consists of the access services network (ASN) and connectivity services network (CSN). The core elements in the ASN are the base station (BS) and ASN gateway (ASN-GW), which are connected over an IP cloud. The functionality across the ASN-GW and BS is split and signaled via R6. The ASN-GW provides security anchoring, traffic accounting, and mobility anchoring (and proxy) for the mobile station (MS). The Mobile IP home agent (HA) in the CSN is used as a global mobility anchor, and Figure 8. DSRC spectrum band 8 is an optional element depending on deployment choices. In the simplified form (Simple IP), the user traffic bypasses the HA in the CSN. The user traffic is tunneled as payload between the BS and the ASN-GW. The Proxy Mobile IP protocol handles mobility between the ASN-GW and the HA. A WiMAX BS can potentially connect to any ASN-GW that it can reach via IP connectivity (flex R6). This flexibility helps decrease mobilityrelated signaling in the network, since the same ASN-GW can serve the user’s active IP session while the user is moving across several different BSs (e.g., ASN-GW relocation is rarely required). The R8 interface can facilitate the context transfer and handover optimization when the user moves from one BS to another. Figure 9 shows the end-to-end Mobile WiMAX network architecture. The WiMAX Forum has defined an architecture that determines how a WiMAX network connects with other networks, and a variety of other aspects of operating such a network, including address allocation and authentication. Figure 10 shows an overview of the architecture. • • • • • • • SS/MS: Subscriber Station/Mobile Station BS: Base station, part of the ASN ASN-GW: The ASN Gateway, part of the ASN CSN: The Connectivity Service Network HA: Home Agent, part of the CSN AAA: AAA Server, part of the CSN NAP: A Network Access Provider Introduction of Vehicular Network Architectures Figure 9. Mobile WiMAX network architecture neXT-generaTion Vehicular neTwork archiTecTure This section discusses the architecture of the Next-Generation Vehicle Network, which is under development at the University of Detroit Mercy by several researchers in collaboration (Mahfoud et al., 2008). in-Vehicle personal computer Automotive manufacturers are planning to provide new Internet and entertainment services inside vehicles. These services need to monitor and collect data collection from increasing the requirement for computing capabilities onboard vehicles (InCode Telecom Group, 2009; Microsoft drives, 2004; Mosra et al., 2004). The largest computer hardware and software companies are now developing suitable platforms that would provide the desired telematics and Internet services for vehicles. For instance, the Intel Corporation has announced plans to bring PCs to cars in a project called Connected Car PC Technology, and has indicated it at many events, including the Consumer Electronics Show in Las Vegas in 1998 (Machen, 2009), where Intel demonstrated a Ford Expedition loaded with a Pentium MMX-based computer. This computer provides voice- activated satellite-based navigation, Internet access, cellular phone calls, games and DVD movies. The IBM infotainment system Imaj provides an example of on-board computers. Imaj comprises of three 12-inch LCD screens built into the car seats facing three rear-seat passengers. These screens provide access to radio, audio player, video player, fax, phone, e-mail, calculator, satellite mapping, local information, word processing, web brows- 9 Introduction of Vehicular Network Architectures Figure 10. The WiMAX overview of the architecture ing, and calendar functions all controlled either by mouse or by speech using IBM via-Voice technology (Ibm.com). Software platforms are being developed for in-vehicle computers to provide voice-activated commands. For instance, Microsoft Corporation has launched the .NET Connected Car (Microsoft drives, 2004), which provides the next generation of software for the melding of computing and communications in the vehicle (Microsoft. com, 2009). Microsoft describes Connected Car as follows: “Tomorrow’s driver can expect better information through systems built on the Microsoft .NET Connected Car software, a new generation of technology that connects individual cars to other information systems through the Internet. Your next car may tell you that you should change the oil sooner than expected. It may also direct you to the closest dealer, help schedule an appointment, notify you of oil change specials, and keep you informed on up-to-date service alerts.” (Microsoft drives, 2004). 10 The .NET Connected Car uses the Windows Automotive operating system, which provides computing services such as voice-activated commands, and video and audio support for entertainment. The .NET Connected Car also includes built-in support for popular wireless technologies, including Wi-Fi, General Packet Radio Service (GPRS), Bluetooth and Code Division Multiple Access (CDMA) (Microsoft drives, 2004). system architecture Figure 11 shows the system architecture, which consists of three subsystems. The first subsystem is the vehicle, which includes the various networks and the on-board PC. Next is the wireless communication network, which provides a telephone service, as well as a WAN (Wide Area Network) connection between the PC and the Internet. The third subsystem is the Internet, which provides a database server for vehicle information. Introduction of Vehicular Network Architectures Figure 11. A block diagram of the next generation vehicle network Vehicle Network and the On-board PC This architecture considers that all the intelligent sensors, actuators and other control units inside the vehicle are implemented by CAN-capable microcontrollers. In this case, the CAN protocol is implemented for networking these devices together. Networking all these devices together in one network would enable them to exchange data among themselves, while also allowing any device to transmit data to an external network, or be accessed from it to provide features such as remote diagnostics, monitoring, data collection and remote firmware upgrade. This architecture makes various hardware and software changes to the vehicle CAN network in order to make this network Internet-enabled. The remote firmware upgrade feature of the CAN nodes requires the most changes. The discussed architecture requires the CAN modules to have Flash memory for code, and an on-module or on-chip (built into the micro-controller unit (MCU)) Flash programming voltage generator. The firmware of the CAN nodes have a specially-developed Bootloader core in the nonvolatile memory of the MCU, thus enabling access to the MCU via the CAN bus through the CAN port of the MCU to reprogram the Flash memory when required. This new feature is essential for facilitating and easing the debugging and reprogramming of the CAN nodes, as well as for enabling this task to be performed remotely over the Internet. The CAN nodes should also be able to operate in Wakeup mode, the nodes would consume the minimum amount of battery power when the vehicle is stopped, while still being able to be woken up remotely, if required, to perform a diagnosis task. To reduce the cost and size, and to minimize technology obsolescence, an on-board computer generally does not have very high storage and processing capacity (Encyclopedia.com, 2009). Additionally, this computer mostly acts as a client depending on the remote server on the Internet, and therefore does not require very high performance. 11 Introduction of Vehicular Network Architectures The In-vehicle, or on-board, PC plays several roles in this architecture. It is the master control unit for the vehicle, the TCU (Telematics Control Unit) (Ibm.com) and a gateway for the different networks inside the vehicle. The gateway role provides connectivity among the in-vehicle networks themselves, and between each of them and the Internet as necessary. Therefore, the PC requires an interface to every network that needs to be connected to it. Additionally, the PC should also have an interface to the Internet to provide connectivity (gateway) between the vehicle network and the Internet, as well as to provide regular Internet access for browsing and entertainment. The Wireless Network The wireless network in this architecture exists to provide communication, phone calls and Internet access inside the vehicle. In this subsystem, the on-board PC is connected to the Internet via a WAN (Wide Area Network) Connection. This WAN is a wireless Internet connection provided by any appropriate wireless communications carrier for cellular phones and similar devices. Speed and availability are the main factors to determine the suitability of that carrier. The GPRS (General Packet Radio Service) (Gsmworld.com, 2009), which is a packet-switching, non-voice service, can be implemented as the wireless Internet connection. It allows information to be sent and received across a mobile telephone network, and provides actual packet radio access and timedivision multiple access (TDMA) to users. GPRS is currently adopted to gain wireless access to many Internet applications and services wirelessly such as chat, textual and visual information including the download of audio, still and moving images; web browsing; email; vehicle positioning with collaboration from GPS; remote LAN access; and file transfer. These applications can be accessed by several GPRS devices (terminals), such as web-enabled cellular phone 12 sets and the wireless PCMCIA cards for laptops and pocket PCs (Gsmworld.com, 2009). GPRS has a theoretical maximum speed of 171.2 kilobits per second (kbps) when using all eight timeslots at the same time. The most important feature of the GPRS is immediacy, which a dialup connection is not required. Users can access the Internet immediately, since they are always connected due to the packet-switching technology adopted by GPRS on the existing circuit-switching GSM (Gsmworld.com, 2009). The Internet An Internet server is required to store uploaded data from the vehicles. Each vehicle has its own account on such server. The manufacturer and the service center can access to the vehicle’s date in that server. The Internet is also adopted to access the computer on board the vehicle and, consequently, to access the vehicle network that is connected to and monitored by that computer. Accessing the on-board computer from the Internet requires the implementation of strict security measures that ensure appropriate access for authorized parties only. Vehicle data can be protected by implementing frameworks for data protection based on privacy and security technologies (Duri et al., 2002). reFerences Duri, S., Gruteser, M., Liu, X., Moskowitz, P., Perez, R., Singh, M., & Tang, J. M. (2002). Framework for security and privacy in automotive telematics. In Mobile Computing and Networking, (pp. 25-32). Eichler, S. (2007). Performance evaluation of the IEEE 802.11p WAVE communication standard. In Proceedings of IEEE Vehicular Technology Conference, (pp. 2199-2203). Introduction of Vehicular Network Architectures Encyclopedia.com. (2009). Telematics is not a question of if, but when. Retrieved from http:// www.encyclopedia.com/doc/1G1-94875067. html FlexRay Communications System Protocol Specification v2.1 Revision A. (2005). FlexRay Consortium. (2009). Retrieved from http://www.flexray.com Gsmworld.com. (2009). Retrieved from http:// www.gsmworld.com/technology/gprs/index. shtml Ibm.com. (n.d.). Focus on the road ahead: IBM puts practical telematics within reach. Retrieved from http://www-306.ibm.com/software/pervasive/info/Telematics_within_reach_050404.pdf IEEE802.16-2005. (2006). Part 16: Air interface for fixed and mobile broadband wireless access systems amendment 2: Physical and medium access control layers for combined fixed and mobile operation in licensed bands and corrigendum 1. IEEE P802.11p/D3.0. (2007). Draft amendment for wireless access in vehicular environments (WAVE). InCode Telecom Group Inc. (2009). Telematics: How economic and technological forces will shape the industry in the U.S. Retrieved from http://www.incodewireless.com/pdfMailer/files/ Telematics_Position_Paper_v11.pdf Jiang, D., & Delgrossi, L. (2008). IEEE 802.11p: Towards an international standard for wireless access in vehicular environments. In IEEE Vehicular Technology Conference, (pp. 2036-2040). Johansson, K. H., Torngren, M., & Nielsen. L. (2003). Vehicle applications of controller area network. Technical Report Department of Signals, Sensors and Systems, Royal Institute of Technology, Stockholm, Sweden, Department of Electrical Engineering, Linkoping University, Sweden. Kim, J. H., Seo, S. H., & Moon, T. Y. (2007). A method of improving the reliability of the gateway system by using OSEK/VDX. In Proceedings of International Conference on Control, Automation and System, (pp. 2328-2833). Machen. L. (2009). Technical marketing engineer, “Intel drives in-vehicle solutions” handheld components division. Retrieved from http:// www.intel.com/technology/magazine/computing/ it04001.pdf Mahfoud, M., Al-Holou, N., & Baroody, R. (2008). Next generation vehicle network: Web enable. In Proceedings of 3rd International Conference on Information and Communication Technologies: From Theory to Applications, (pp. 1-7). Microsoft drives to Las Vegas. (2004). Connected Concept Cars. Retrieved from http://www.windowsfordevices.com/news/NS5155857065.html Microsoft.com. (2009). Windows Automotive. Retrieved from http://www.microsoft.com/windowsautomotive/wa5/default.mspx. Mosra, S. R., Shanker, S., & Mahmud, S. M. (2004). An intelligent architecture for issuing intersection collision warnings (pp. 176-183). National Defense Industries Association (NDIA). OSEK VEX Portal, (n.d.). Retrieved from http:// www.osek-vdx.org Pazul. K. (1999). An introduction to the CAN protocol that discusses the basics and key features. Microchip Application Note #AN713. Robert Bosch Gmb, H. Stuttgart, Germany, (1991). CAN Specification ver. 2.0. Sun, Y., Wang, F. Y., Wang, Z. X., Qiao, X., & Wang, K. F. (2006). A scheduling algorithm for vehicular application specific embedded operating systems. In Proceedings of IEEE International Conference on Systems, Man and Cybernetics, (pp. 2535-2540). 13 Introduction of Vehicular Network Architectures Wu, H., Fujimoto, R., Hunter, M., & Guensler, R. (2005). An architecture study of infrastructurebased vehicular networks. In ACM MSWiM, Montreal, Canada, (pp.36-39). Xiang, W., Richardson, P., & Guo, J. (2006). Introduction and preliminary experimental results of wireless access for vehicular environments (WAVE) systems. In Proceedings of 3rd Annual International Conference on Mobile and Ubiquitous Systems, (pp.1-8). 14 Xu, Y. N., Kim, Y. E., Cho, K. J., Chung, J. G., & Lim, M. S. (2008). Implementation of FLexRay communication controller protocol with application to a robot system. In Proceedings of 15ht International Conference on Electronics, Circuits and Systems, (pp. 994-997). 15 Chapter 2 Introduction of Vehicular Network Applications Yao-Chung Chang National Taitung University, Taiwan, R.O.C. absTracT Information and Communication Technology (ICT) is concerned with all the technologies that manage, process, and communicate information. It is also named as telematics, combining two words: telecommunications and informatics, which is widely used in the application of Global Positioning System technology integrated with computers and in the mobile communications technology for automotive navigation systems. Table 2.1 and Table 2.2 respectively list the telemetric applications from user’s point of view and the practical applications of vehicular telematics. Four applications of the vehicular network are discussed in this chapter. The first section introduces the vehicular network application services. The second section discusses the vehicular network application management. The third section provides the platform technologies of vehicular network application. Finally, future vehicular network application and deployments are presented in the fourth section. inTroducTion Service bundle provision systems (Choi et al., 2005; Han et al., 2005) enable users to search, download and install service applications that can be operated on a user terminal in an open telematics environment (shown in Figure 1). The system has incorporated the telematics gateway in the service providers including the wireless optimized TCP, the wireless optimized HTTP, the SMS gateway, the push module and the framework, as illustrated in Figure 1. Such a system has three main components: gateway, framework and world telematics protocol. • Gateway: The gateway allows developers to write a telematics server application that DOI: 10.4018/978-1-60566-840-6.ch002 Copyright © 2010, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global is prohibited. Introduction of Vehicular Network Applications Figure 1. Open telematics service • • can be operated irrespective of the gateway of the mobile network. Framework: The developers can write applications, without knowing about the details of integrating the related servers distributed in networks, by utilizing APIs supported by the framework. World Telematics Protocol (WTP) (WTP1.0 Specification, 2004): WTP defines a protocol to exchange messages between a telematics terminal in a vehicle and a telematics service center. Telematics service developers and service providers can develop and provide telematics services that do not depend on devices and service carriers. See Tables 1 and 2. Table 1. Telematics applications from user’s point of view Audio (CD / Radio) 16 Telephone The telematics portal framework (shown in Figure 2) is mainly divided into three components. The first component is a provisioning service part that provides a telematics terminal with telematics service bundles. The second component is a service bundle management module that provides functionality for deploying and managing service bundles. The third component is a repository module that not only supports other two components and stores service bundles and related data, but also allows search and manage service bundles. Additionally, it supports various software providing protocols including JNLP, MIDP and J2EE client provisioning. Table 2. Practical applications of vehicle telematics Vehicle tracking Wireless vehicle safety communications Trailer tracking Emergency warning system for vehicles Cold store freight logistics Intelligent vehicle technologies Telephone Navigation Fleet management Car clubs Video Speech recognition Satellite navigation Auto insurance Short messaging (SMS) User interface for body electronic Mobile data and mobile television Introduction of Vehicular Network Applications Proactive services (Gura et al., 2001) describe a mobile application for traffic telematics based on the JINI middleware. The main contribution proactively is the definition of a user agent that closely interacts with a service discovery and lookup facility. The lookup service is available in each vehicle and provides a set of both user-level services and lower-level services for accumulating, improving and distributing information obtaind from various embedded sensors. The low-level services form the basis for context creation. To exploit these services proactively, a user agent is designed to perform the following tasks: • • User agents monitor available services on other vehicles. They are parameterized by user profiles, and are selected based on user settings. User agents assess the lower-level floating car data to generate the context based on which decisions involve services. • User agents provide a user-defined appearance of user-level services. Thus, a single look-and-feel interface encompasses the various interfaces of various services as much as possible. The Road-Look-Ahead (RLA) Service (Gura et al. 2001) is composed of all relevant elements of a proactive service listed in the above three statements. This service dynamically links a car, e.g. a truck as shown in Figure 3, advertising an image from the local telematics system. To select an image that may be offered by many cars in reach, the user agent needs to perform a selection according to the context information, including the speed of the car and the direction, thus filtering out cars that are far away or come from the opposite direction. The system architecture is based on Java (Sun Microsystems Inc., 2000) and JINI (Sun Microsystems Inc., 1999) technologies. The concept of a service is essential to the JINI system and gives the system various advantages. Figure 2. The portal framework overview 17 Introduction of Vehicular Network Applications First, the low-level context information such as position, direction and speed can be abstracted and encapsulated in a JINI location service that abstracts from all specific details of the navigation devices. Second, JINI provides several mechanisms linked to mobility. Growing consumer demand for access information in remote, mobile environments has sparked interests in In-Vehicle Telematics Systems (IVIS). This work introduces the EmergeITS project (Reilly et al., 2002), which focuses on using IVTS in emergency fire services. In particular, this work describes a distributed service-based architecture, based on the JINI middleware technology, which can be employed to provide fault-tolerant application services to remote in-vehicle computers and mobile devices including Palm devices and WAP phones. EmergeITS adopts the JINI middleware (Figure 4) to provide a service-based architecture that can manage, configure and provide application services to remote in-vehicle computers and Palm devices. Figure 3. Road-look-ahead service 18 Vehicular neTwork applicaTion managemenT Communications resources management (Punnoose et al., 2001) for advanced telematics applications is divided into three parts: external communications, internal communications and interface to resource manager. The architecture which supports mechanisms for internal and external communications is shown in Figure 5. Communications between the in-vehicle components and the exterior pass through a router gateway. This routing module selects the appropriate external communication method for data connections according to the parameters such as cost, availability and QoS requirements of the applications. The resource manager interacts with this routing module to advise it of the QoS requirements of applications. Inter-agent communication within the vehicle must be flexible in order to facilitate the addition of new software agents. A model that allows the Introduction of Vehicular Network Applications Figure 4. SunLabs’ concept of car software architecture Figure 5. Management of internal and external communications transfer of data without strict pre-definition (Rajkumar et al., 1995) is preferred, so that all agents that tap into this infrastructure can be publishers or subscribers of information, or both. The information that is being published is transparent to the underlying infrastructure, which merely transports it from the publishers to the subscribers. Publishers post information on logical channels. Other agents can be subscribers to this information. 19 Introduction of Vehicular Network Applications The resource manager can be a single or distributed entity. Client applications communicate with the resource manager via network packets. Therefore, the interface to the resource manager can be embedded in the same processor as the client application or in a different machine within the vehicle. The only information required by the client is the location of the resource manager. A modern vehicle contains many electronic devices (Nolte et al, 2005), linked by Local Interconnect Network (LIN), Controller Area Network (CAN) and FlexRay (FlexRay Communications System Protocol Specification v2.1 Revision A, 2005), to fulfill the needs of customers and to enhance the performance of the vehicle. The smart vehicle management system (SVMS) (Seo et al., 2007) is composed of a gateway, handset and vehicle management program (VMP). The increasing number of electronic devices enables automatic control and management of vehicles. In particular, automatic management of a vehicle provides significant benefits to owners (driv- ers) by decreasing maintenance and increasing convenience. The proposed SVMS adopts CDMA wireless communication (Chan et al., 2003) to control and manage a vehicle (Bjic et al., 2002). CDMA allows vehicle owners to access their vehicles anywhere anytime. Additionally, a handset can be adopted as the remote-controller for the vehicle. VMP provides a graphic user interface that manages the vehicle and owner. The gateway plays in a relay role. The developed gateway is an intermediate among handsets, VMPs and vehicular devices, since this system supports LIN, CAN, FlexRay and CDMA interfaces. Thus, the proposed SVMS not only achieves telematics and a ubiquitous environment within a vehicle, but also manages electrical components. Figure 6 illustrates the VMP running on a PC with a RS-232 interface. The CDMA modem is linked to the PC via RS-232. The CDMA modem transfers the received messages to the VMP, and the commands sent to the gateway. The handset Figure 6. Smart vehicle management system prototype design 20 Introduction of Vehicular Network Applications accesses the gateway directly by using the SMS. The gateway is linked with three embedded systems. For instance, embedded system #3 operates several LEDs based on VMP commands or handset messages. Each embedded system is connected using CAN and LIN. Additionally, all operations of each process are printed in VMP and saved in a file by a log writer. The application presentation of telematics is no restricted to the classical embedded control system, but instead it covers a broad range from driver assistance to infortainment and vehicle information management. Management of various telematic applications requires new application managers (Kim et al., 2006) for the automotive domain to support the specific software components, corresponding to the vehicle status, and eventually combine the system parts to form a single reliable and manageable system. This work develops a novel architecture for application management in in-vehicle software (shown in Figure 7). This architecture is based on international standards, including OSGi (Open Service Gateway Initiative, 2003) framework (two major platforms are Prosyst’s mBedded Server (Prosyst, 2009) and Gatespace’s e-Services Platform (Gatespace, 2009) and AMI-C Specification (AMI-C, 2002). Telematics technology, which adopts various telecommunications such as DMB, CDMA, WRAN and DSRC, is increasingly used to provide services for in-vehicle telematics systems or context aware automotive telematics (Zhang et al., 2004). This service bundles provide in-vehicle applications such as vehicle status monitoring service, vehicle tracking and local advertising, can be realized using. This architecture allows in-vehicle terminals to provide various telematics services to improve safety of drivers. Vehicular neTwork applicaTion plaTForm Figure 8 shows the architecture and design of an OSEK/VDX (O’Donnell, 2003) (includes specifications for OSEK Operating System (OS) (OSEK Group, 2005), OSEK Communication (COM) (OSEK Group, 2004), OSEK Network Manager (NM) (OSEK Group, 2004) and OSEK Implementation Language (OIL) (Zheng et al., 2004)) and OSGi-based Embedded Software Platform (Sun et al., 2007). Three components are introduced: The Figure 7. Architecture of application manager 21 Introduction of Vehicular Network Applications remote vehicular service platform component has high computation ability and huge data storage. The vehicle/home interactive platform component is another OSGi-based framework gateway. The vehicle driver can interact with it via the wireless connection and manage home/office appliances while driving. The roadside system component refers to ITS infrastructure such as intelligent traffic light controllers and communication base stations. Figure 9 shows OSGi-based remote vehicular service platform (Wang et al. 2004). The vehicular integration services mainly combine diagnostics and prognostics to form the optimal vehicular control algorithms download services. These services are generally provided by automotive manufacturers or third party developers. Vehicular correlation services mainly provide facilities such as GPS location and information transmission management. Transferring information accurately in real-time is a key to the quality of service. Individual information services provide services including entertainment search and download, information service and vehicle/home interactive service. Telematics open portal applications zone (TOPAZ) (Lee et al. 2005) is under development by the IBM Ubiquitous Lab. Figure 10 describes two telematic applications based on TOPAZ platform (Choi et al., 2006). These applications are a call-taxi service and an emergency rescue service for patient in u-health scenarios. It builds three Figure 8. Architecture of the OSEK/VDX and OSGi-based embedded software platform 22 Introduction of Vehicular Network Applications Figure 9. OSGi-based remote vehicular service platform different applications end-to-end within a short time period and minimal manpower. Call taxi application provides the nearest available taxi to a customer through vehicle tracking information until the assigned taxi arrives at the customer’s location. The main parts of the application service are service clients, call center server and vehicle dispatch server (VDS). The call center server does not directly connect to TOPAZ servers, but instead interacts with VDS and a customer to transfer the requested data between them. Emergency rescue service for patient is a preventive service that monitors biological status of patient using ubiquitous sensors attached to the patient body, predicts the critical situation of the patient, and dispatches the emergency vehicle to the patient’s location. The Health management server (HMS) was developed to interact with patients’ devices using the same VDS adopted in the call taxi application, but to dispatch emergency vehicles rather than taxis. The proposed emergency rescue service application adopts the spatiotemporal event detection environment STEDE (Munson et al., 2005; Lee et al., 2005) rule-based engine for event detection in TOPAZ to determine the critical status of the patient. The growth in Internet subscribers has accelerated at an exponential pace. Wired Internet access serves subscribers only at attached points, while wireless communications provide enjoy ubiquitous/pervasive Internet access. Recent advances in wireless inter-vehicle communication systems have generated major opportunities for the deployment of a wide variety of services to vehicles (Dikajakos et al., 2008). Telematics provide innovation and new technology in the automotive sector, while significantly improving the driving experience (Grymek et al., 2007). This investigation combines wireless services with human-computer interaction to provide a ubiquitous/pervasive computing environment for telematics services. IETF defines mobile IP facilities to provide roaming methods in wireless computing. The NEMO (NEtwork MObility) (Emst, 2006; Devarapalli et al., 2005) working group has improved roaming schemes to enable subscribers to use mobile routers for data transmission without worrying about service environments (Tseng et al., 2007). 23 Introduction of Vehicular Network Applications Figure 10. Application deployment in TOPAZ A mobile router can be installed on any vehicle. The vehicular router communicates with external vehicular networks. Telematics application users within a moving vehicle do not need to perform handoffs individually. The entire vehicle is a subnetwork and performs handoff procedures using the installed vehicular router. Figure 11 illustrates the Mobile telematics computing environment. This study develops an embedded WiMAXbased network mobility system on an Intel IXP425 network processor platform to deploy vehicular routers for vehicular network telematics computing (Chen et al., 2009). A mobile router can be installed on any vehicle. The vehicular router is responsible for communicating with external vehicular networks. When a vehicle moves, telematics users within the vehicle do not need to perform handoffs individually. The whole vehicle is a sub-network and performs handoff procedures using the installed vehicular router. The concept of mobile telematics computing environment is shown in Figure 11. The mobile IPv6 scheme, which includes packet formatting, 24 receiving, sending and processing, is established in the vehicular router. This work also enhances mobile IPv6 technology and network mobility. FuTure Vehicular neTwork applicaTion and deploymenT A client-server application and Internet based application were developed. Users performed various remote experiments on a mobile robot, focusing on motor control, obstacle avoidance and image processing, and applying these features to trajectory control using the Common Gateway Interface (CGI) to access the robots (Backes et al., 2000; Schilling, 2001). The remote user is connected via the Internet to a dedicated computer, which controls and monitors the mobile robot (Popescu et al. 2008). The client was programmed in Java, while server was programmed in Visual Basic 6. The Visual Basic server application handles the application administration module, which involves user management, and communication with the Introduction of Vehicular Network Applications Figure 11. Mobile telematics computing control board and microcontroller program editor. An AXIS 205 network video camera sends live video of the mobile robot to the user’s browser. The Internet provides a major opportunity for robotics education by enabling the development of remote laboratories, where students can gain practical experience deploying hardware resources efficiently. The system reduces the time and space constraints normally associated with the traditional laboratory, thus making equipment available to more students. The InternetCAR project (Ernst et al., 2003; InternetCAR Project, 2003) develops the architecture needed to connect a car to the Internet, but does not address the problems of infrastructure to enable the services. Telematics functions were configured on a dedicated device, commonly called a middlebox (Srisuresh et al., 2002). These middleboxes adopt programmable network processors in order to attain the required processing and forwarding speeds, while communicating with each other and the back-end server using standard middleware components. Urban traffic prediction with floating car data is an example of emerging telematic applications that exploits of global networking capabilities (Chen et al., 2004). A middlebox is a device performing a service that needs application logic, but is executed in the network (shown in Figure 13). The aggregation function receives UDP packets, processes them as described above, and generates asynchronous XML/RPC messages. The function then sends these messages to the server. XML/RPC is an RPC protocol that utilizes HTTP as the transport protocol and XML for data representation. The front-end server was implemented as a Servlet running within the WebSphere Servlet engine. A Mobile Agent-based Middleware for vehicle telematics is proposed in (Guo et al., 2007). The JNomad, a framework that integrates JINI technology with mobile agents, is developed to meet these deployment challenges. JINI is a novel technology that considers the pervasive, ubiquitous, and dynamic distributed computing requirements within a single architecture. To address the mobility and frequent network partition 25 Introduction of Vehicular Network Applications Figure 12. Telematics architecture for mobile robot Figure 13. Overview of middleboxes problem, this investigation designed and implemented a Code-on-Demand Service Model using JINI technology, and developed a Mobile-Agent based Service Model. JNomad is a mobile agent-based framework consisting of three major components. The first component is composed of mobile agents, i.e., entities that are expected to perform various tasks. The second component contains the mobile agent hosts, which provide the platform executed by the mobile agents. The third component comprises lookup and location services that provide dynamic service discovery and location updates. A typical 26 sequence of interactions among different components within the framework is shown in Figure 14. Using the JINI architecture, virtually any service may interact freely in a network without the requirement for complex protocols, messaging drivers, operating systems or cabling. Security issues are also very important topics in vehicular applications. A general architecture for a trusted context provider is proposed in (MOSQUITO Project, 2005). Different security mechanisms can be used in different situations in this architecture. Information Connector is presented for vehicular environments. A general Introduction of Vehicular Network Applications Figure 14. The agent-based framework using JINI technologies Figure 15. Vehicular security architecture 27 Introduction of Vehicular Network Applications architecture for security in vehicular environment is presented in (Nowey et al., 2006). Figure 15 illustrates the overall architecture of vehicular security architecture (Nowey et al. 2007), including Service Plane, Security Middleware and Context Aware Applications. • • • Service plane: The security layer is configured using the security manager, which provides the appropriate interface. Additionally, the security manager sets the address of the position based routing entity, thus protecting the location privacy of the users. Security middleware: The three major parts of the security middleware are trust establishment services, revocation services and privacy services. Context aware applications: Contextaware applications perform actions such as issuing a warning to the driver or peer applications based on the context information. BMW, in collaboration with Telematics Service Providers (TSPs) Connexis (Connexis, 2009) and WirelessCar (WirelessCar, 2009), has developed a new telematics framework and a technologyneutral telematics protocol to bring new flexibility and scalability to the automotive sector called Next Generation Telematics Protocol (Next Generation Telematics Protocol, 2009). Vehicle manufacturers have historically provided proprietary services to customers, and have been dependent on a single TSP for delivery of these services in a specific market. This supply chain inflexibility has created difficulties for providers in gaining economies of scale and promoting their products and services. Although an open and standardized approach to delivering services has clear benefits for the marketplace, previous standardization efforts have encountered barriers to adoption, because they have focused on replacing existing protocols instead of integrating them. The proliferation of new technologies suggests that future in-vehicle 28 devices are likely to access services using multiple methods and technologies. reFerences AMI-C. (2002). Software API SpecificationsCORE APIs. retrieved September 20 2009, from http://www.ami-c.org Backes, P., Tso, K., Norris, J., Tharp, G., Slostad, J., Bonitz, R., & All, K. (2000). Internet based Operations for the Mars Polar Lander Mission. In The 2000 IEEE International Conference on Robotics & Automation, (pp. 2025-2032). Bjic, E., & Chaxel, F. (2002). Auto-ID Mobile Information System for Vehicle Life Cycle Data Management. IEEE Systems . Man and Cybernetics, 4, 6. Chan, Alex Y. M., & Lu, W. P. (2003). Architecture for Wireless Access in Vehicles. IEEE Vehicular Technology Conference, 5, 3336-3340. Chen, A., Jain, N., Perinola, A., Pietraszek, T., Rooney, S., & Scotton, P. (2004). Scaling RealTime Telematics Applications using Programmable Middleboxes: A Case Study in Traffic Prediction. In First IEEE Consumer Communications and Networking Conference, (pp. 388-393). Chen, J. L., Chang, Y. C., Lin, Y. S., & Du, H. W. (2009). Embedded WiMAX-based Vehicular Router for Telematics Computing. In The 11th International Conference on Advanced Communication Technology, (pp. 1857-1862). Choi, D. H., Ro, S. H., Lee, K. S., Lee, W. N., Choi, S. Y., & Pae, Y. W. (2006). Telematics Applications Based On TOPAZ Platform. In IEEE 6th International Conference on Computer and Information Technology, (pp.262-268). Introduction of Vehicular Network Applications Choi, J. W., Han, W. Y., Kim, C. S., & Kwon, O. C. (2005). Open Telematics Services Deployment on the Gateway and Framework Independent on Mobile Networks. In International Conference on Wireless Networks, (pp.374-379). Connexis website. (n.d.). Retrieved September 20, 2009, from http://www.connexis.com/ Devarapalli, V., Wakikawa, R., Petrescu, A., Thubert, P. (2005). Network Mobility (NEMO) Basic Support Protocol RFC 3963. Dikajakos, M. D., Florides, A., Nadeem, T., & Iftode, L. (2008). Location-Aware Services over Vehicular Ad-Hoc Networks using Car-to-Car Communications. IEEE Journal on Selected Areas in Communications, 25(8), 1590–1602. doi:10.1109/JSAC.2007.071008 Emst, T. (2006). Network Mobility Support Goals and Requirements. Ernst, T., Uehara, K., & Mitsuya, K. (2003). Network Mobility from the InternetCAR Perspective. In 17th International Conference on Advanced Information Networking and Applications, (pp. 19–26). FlexRay Communications System Protocol Specification v2.1 Revision A, (2005). Gatespace website. (n.d.). Retrieved September 20 2009, from http://www.gatespacetelematics.com/ Gerlach, M., & Fokus, F. (2007). Trust for Vehicular Applications. In 8th International Symposium on Autonomous Decentralized Systems, (pp. 295-304). Grymek, L., Singh, S., & Pattipati, K. (2007). Vehicular Dependence Adds to Telematics’ Allure. IEEE Potentials, 26(2), 12–16. doi:10.1109/ MP.2007.343053 Guo, J. H., & Xing, G. M. (2007). Using Mobile Agent-based Middleware to Support Distributed Coordination for Vehicle Telematics. In 21st International Conference on Advanced Information Networking and Applications Workshops, (Vol. 2, pp.374-379). Gura, N., Held, A., & Kaiser, J. (2001). Proactive Services in a Distributed Traffic Telematics Application. Workshop on Mobile communication over Wireless LAN: Research and Applications. Retrieved from http://www.informatik.uni-ulm. de/rs/projekte/core/ Han, W. Y., Kwon, O. C., Park, J. H., & Kang, J. H. (2005). A Gateway and Framework for Interoperable Telematics Systems Independent on Mobile Networks. ETRI Journal, 27(1), 106–109. doi:10.4218/etrij.05.0204.0038 Internet, C. A. R. Project (2003). InternetCAR webpage at WIDE. Retrieved September 20, 2009 from http://www.sfc.wide.ad.jp/InternetCAR Kim, M. J., Choi, Y. J., Moon, Y. B., Kim, S. J., & Kwon, O. C. (2006). Design and Implementation of Status based Application Manager for Telematics. IEEE 8th International Conference on Advanced Communication Technology, (pp. 20-22). Lee, S. W., Munson, J., Lee, D. R., Thompson, G., & Park, J. (2005). A Rule-based System for Sense-and-Respond Telematics Applications. In The First Korea/Japan Joint Workshop on Ubiquitous Computing & Networking Systems, (pp 109-114). Lee, W., Tak, Y., Byun, W. H., & Pae, Y. W. (2005). Toward an Open Telematics Infrastructure. IPSJ SIG Technical Reports, (pp. 443-448). MOSQUITO Project. (2005). Mosquito Architecture. White Paper. Retrieved September 20, 2009 from http://www.mosquito-online.org/ 29 Introduction of Vehicular Network Applications Munson, J., Lee, S. W., Lee, D. R., Wood, D., Thompson, G., & Cole, A. (2005). A Rule-based System for Sense-and-Respond Telematics Services. Workshop on End-to-end, Sense-andRespond Systems, Applications and Services, (pp. 31-36). Next Generation Telematics Protocol website. (n.d.). Retrieved September 20 2009, from http:// www.ngtp.org Nolte, T., Hansson, H., & Bello, L. L. (2005). Automotive Communication – Past, Current and Future. In 10th IEEE Conference on Emerging Technologies and Factory Automation, 1, 985992. Nowey, K. P. T., & Mletzko, C. (2006). Towards a Security Architecture for Vehicular ad hoc Networks. In First International Conference on Availability, Reliability and Security, (pp. 374-381). O’Donnell, M. (2003). OSEK/VDX: Driving the Standard for Optimized Embedded Systems. Embedded Computing Design. Open Service Gateway Initiative. (2003). OSGi Specification Release 3. Retrieved September 20 2009, from http://www.osgi.org OSEK Group (2004). OSEK/VDX Network Management Concept and Application Programming Interface, Version 2.5.3. OSEK Group (2004). OSEK Implementation Language, Version 2.5. OSEK Group (2005). OSEK/VDX Operation System, Version 2.2.3. Popescu, D., & Selisteanu, D. (2008). Web based Telematics Application for Robotics. In the 3rd International Multi-Conference on Computing in the Global Information Technology, (pp. 19-24). Prosyst website. (n.d.). Retrieved September 20 2009, from http://www.prosyst.com/ 30 Punnoose, J., Tseng, S., Wang, S., Nikitin, V., & Schlesinger, T. E. (2001). Communications Resources Management for Advanced Telematics Applications. Transportation Society Conference. Rajkumar, R., Gagliardi, M., & Sha, L. (1995). The Real-Time Publisher/Subscriber Inter-Process Communication Model for Distributed Real-Time Systems: Design and Implementation. Real-Time Technology and Applications Symposium, (pp. 66–75). Reilly, D., & Taleb-Bendiab, A. (2002). A ServiceBased Architecture for In-Vehicle Telematics Systems. In the 22nd International Conference on Distributed Computing Systems Workshops, (pp.741-742). Schilling, K. (2001). Virtual Laboratories for Mobile Robots. IFAC Workshop on Internet Based Control Education, (pp. 115-120). Seo, S. H., Moon, T. Y., & Kim, J. H. (2007). Smart Vehicle Management System by using Gateway, Hand-set and VMP. International Conference on Control, Automation and Systems, (pp.17-20). Srisuresh, P., Kuthan, J., Rosenberg, J., Molitor, A., & Rayhan, A. (2002). Middlebox Communication Architecture and Framework. IETF Internet Draft, retrieved September 20 2009 from http:// www.jdrosen.net/papers/rfc3303.txt Sun, Y., Huang, W. L., Tang, S. M., Qiao, X., & Wang, F. Y. (2007). Design of an OSEK/VDX and OSGi-based Embedded Software Platform for Vehicular Applications. IEEE International Conference on Vehicular Electronics and Safety, (pp. 1-6). Sun Microsystems Inc. (1999, November). JINI Specification v1.0.1. Retrieved from http//:www. sun.com/JINI/ Introduction of Vehicular Network Applications Sun Microsystems Inc. (2000). Java Media Framework API Guide, 2000. Retrieved from http://java. sun.com/products/java-media/jmf/ Tseng, Y. C., Chen, J. J., & Cheng, Y. L. (2007). Design and Implementation of a SIP-based Mobile and Vehicular Wireless Network with PUSH Mechanism. IEEE Transactions on Vehicular Technology, 56(6), 3408–3420. doi:10.1109/ TVT.2007.906986 Wang, F. Y., Li, J. X., & Liu, X. J. (2004). OSEK/ VDX and OSGi-based Embedded Platform for Vehicular Software and its Key Techniques. Chinese High Technology Letters, (pp.131-136). WirelessCar website (n.d.). Retrieved September 20 2009, from http://www.wirelesscar.com/ WTP1. 0 Specification (2004). Electronics and Telecommunications Research Institute (ETRI), Korea. Zhang, D., Wang, X. H., & Hackbarth, K. (2004). OSGi based Service Infrastructure for Context Aware Automotive Telematics. IEEE 59th Vehicular Technology Conference, 5, 2957-2961. Zheng, N. N., Tang, S., Cheng, H., Li, Q., Lai, G., & Wang, F. W. (2004). Toward Intelligent DriverAssistance and Safety Warning Systems. IEEE Intelligent Systems, 19(2), 8–11. doi:10.1109/ MIS.2004.1274904 31 32 Chapter 3 Introduction to ITS and NTCIP Da-Jie Lin Feng Chia University, Taiwan, R.O.C. Chyi-Ren Dow Feng Chia University, Taiwan, R.O.C. absTracT Intelligent Transportation Systems (ITS) combines high technology and improvements in information systems, communication, sensors, and relevant mathematical methods with the conventional world of surface transportation infrastructure to increase the capacity of transportation systems and to improve the level of services. There are four major goals of ITS, including safety, environmental protection, efficiency, and economy. NTCIP (NTCIP Standard 9001, 2002; DISA et al., 1997) is a set of communications protocols and data definition standards designed for various needs of ITS services and applications. The key goals of the NTCIP open-standards effort are interoperability and interchangeability. Interoperability refers to the ability for multiple devices to work together as a single system and interchangeability refers to the ability to use multiple brands of a device on the same communications channel. Accompanying the social and economic development, traffic congestion and delay have become major issues in most areas around the world. How to use readily available technologies to increase the capacity of transportation systems and to improve the level of service has become one of major solutions to solve transportation problems that people are facing. This is the motivation of Intelligent Transportation Systems development. NTCIP is a set of communications protocols and data definition standards designed for various needs of ITS services and applications. These standards are intended to handle these needs in the two areas: Center-to-Field (C2F) and Center-to-Center (C2C) communications. DOI: 10.4018/978-1-60566-840-6.ch003 Copyright © 2010, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global is prohibited. Introduction to ITS and NTCIP inTroducTion Intelligent Transportation Systems (ITS) (ITS Standard et al., 2006; ITE et al., 2003; RITA et al., 2002) combines high technology and improvements in information systems, communication, sensors, and relevant mathematical methods with the conventional world of surface transportation infrastructure. definition of intelligent Transportation systems Intelligent Transportation Systems are defined commonly as follow: systems utilizing newly developed information and communications technology to transportation infrastructure, management systems, and vehicles to improve safety, efficiency, travel time and comforts and reduce vehicle wear, delay, and fuel consumption. Intelligent transportation systems include several systems using different technologies, such as traffic signal control systems; vehicle allocation and navigation systems; container management systems; changeable message signs (CMS); automatic number plate recognition (ANPR) or law enforcement equipments (speed cameras, CCTV, surveillance cameras, ... etc.) to more advanced applications such as traffic control centers that integrate live data and feedback from a number of sources, such as microwave/infrared vehicle detectors, parking guidance and information systems; weather information sensors, analyze all information through different models and then control the signal systems in the facilities to improve the traffic condition. of advanced communication, sensing, and information processing technologies. When integrated into the transportation system’s infrastructure, and into vehicles themselves, these technologies relieve congestion, improve safety, and enhance productivity. Overall, the goal of ITS is to utilize advanced technologies to increase the efficiency of limited transportation resources, to increase user convenience and the living quality. The content of ITS includes technologies and policies. The integration of both technologies and policies and the application of the integration are the core of ITS. There are four major goals of ITS: safety, environmental protection, efficiency, and economy. The details of these major goals are listed in Table 1. A dedicated website, ITS Benefits, Costs and Lessons Learned Databases (http://www. benefitcost.its.dot.gov/) provides updated and comprehensive information about the benefit and cost of ITS. The benefit of ITS can be classified in two application approaches: (1) Intelligent Table 1. Major goals of ITS Goals Objectives 1. Safety 2. reduced incident damages 2. Environmental Protection 1. less air pollution 2. less greenhouse effects 3. less noise 4. less fuel consumption 5.less new facilities & constructions 3. Efficiency 1. less travel time 2. increased capacity goals and benefits of intelligent Transportation systems Intelligent transportation systems provide a set of strategies for the transportation problems related to safety and congestion, while accommodating travel demands of users and freight through the use 1. reduced incident frequency 3. reduced operating costs 4. higher user satisfaction 4. Economy 1. increased industry production 2. more relevant job opportunities (Source:Taiwan logistics management yearbook -Intelligent Transportation Systems, 2001) 33 Introduction to ITS and NTCIP Infrastructure and (2) Intelligent Vehicles. In terms of infrastructure part, the benefit of ITS includes better control in arterial and freeway management, vehicle crash prevention and safety, road weather management and warning systems, transit systems in improved level of service, traffic incident management to reduce any incident impacts, better emergency management, labor- and error-free electronic payment and precise pricing, readily available traveler information to reduce user costs, ...etc. On the other hand, intelligent vehicles increase driving safety through new technologies such as collision avoidance and driver assistances such as car navigation systems and driver monitoring systems. subsystems of intelligent Transportation systems There are commonly discussed nine ITS subsystems that have their own functions and required technologies. They work either jointly or independently, depending on the scenarios. Advanced Traffic Management Systems (ATMS) Advanced Traffic Management Systems is the core and foundation of Intelligent Transportation Systems. ATMS uses technologies such as detecting, sensing, communications, and control to collect the roadside data that will be transmitted back to a traffic control center. Data from different sources will be analyzed and traffic control center will determine their control strategies based on the result of analysis in order to systematically optimize the whole network. The analysis results will be used to control traffic (kight or signau) and the analysis results will be relayed to the public. ATMS increases the efficiency of transportation networks as well. ATMS emphasizes the integration with other systems and real-time traffic data. Related technologies and applications includes traffic signal control, ramp control, automatic incident detections, automatic vehicle locations (AVL), changeable message signs (CMS), Geographic Information Systems (GIS), Weigh-in-Motion (WIM), Automatic Vehicle Classification (AVC), Electronic Toll Collection (ETC), Automatic Vehicle Identification (AVI), …etc. Figure 1. Advanced traffic management systems (Source:ATMS) 34 Introduction to ITS and NTCIP As shown in Figure 1 illustrates the interactions among control center, infrastructure, and vehicles. Advanced Traveler Information Systems (ATIS) Advanced Traveler Information Systems are designed to provide users the information they need to plan their trips. The factor considered in trip planning consists of transportation modes, costs, schedules, accessibilities, connecting transportation, etc.. ATIS utilizes the data collected from different sources and processes these data to make them available and understandable to users. The platform the ATIS uses to deliver the information to users includes TV at home, radio in vehicles, changeable message sign (CMS) on the roadside or in transit stations, internet (Feit et al., 1993) (wired or wireless connection), cellular phones, etc.. Through this real-time, updated travel information, the uncertainties in travel have been greatly reduced and users have better control of their time to increase their productivity (see Figures 2 and 3). Relevant applications and technologies include Highway Advisory Radio (HAR), Global Positioning Systems (GPS) and Geographic Information Systems (GIS), On Board Units (OBU), Wireless communications, and Integrated Service Digital Network (ISDN), etc.. Advanced Public Transportation Systems (APTS) Advanced Public Transportation Systems (APTS) use the technologies of ATMS, ATIS and AVCSS and apply them in the field of public transportation to improve the service quality, efficiency, and attractiveness of existing public transportation services such as buses or trains. Users, instead of being delayed due to the uncertainty of the transit systems, have better utilization of their time. On the other hand, transit agencies have a better control over their own fleet to improve their management. Figure 2. GPS navigation systems - to provide route guidance (Source: Japan ITS HANDBOOK 20062007) 35 Introduction to ITS and NTCIP Figure 3. Changeable message sign (Source: http://www.flickr.com/photos/moogal/442796974/) In the case of buses, dynamic bus information systems, one APTS system for example, provide passengers detailed information such as route information, real-time bus location, estimated bus arrival time, connecting information (other bus routes or modes) so passengers can plan their travel in a more efficient manner. Related applications and technologies for APTS include Automatic Vehicle Monitoring (AVM), Automatic Vehicle Locating (AVL), wireless communications, Electronic Fare Payment (EFP), bus scheduling systems, etc. See Figure 4. Advanced Vehicle Control and Safety Systems (AVCSS) Advanced Vehicle Control and Safety Systems integrate technologies such as sensors, onboard computers, communications, electronics, and control to improve the vehicles and transportation facilities in terms of safety, capacity, and level of service. The advanced vehicle control and safety technologies have been deployed in vehicles to assist drivers to improve their driving skills, to reduce the likelihood of human errors and to increase the safety, especially fatigue has been one of the major 36 reasons of transportation incidents. Increasing the level of automation in driving will influentially reduce the incident possibilities. For example, the onboard driver monitoring systems will monitor the physical condition of the driver when he is executing his driving duty. Another example is collision avoidance system that will work to slow down the vehicle when it is too close to the vehicle in the front and the driver is not responding to this situation accordingly (Figures 5 and 6). Related AVCSS technologies and applications include automatic parking systems, collision avoidance, communications between vehicle and facility, driving behavior monitoring, human factor engineering, etc. Commercial Vehicle Operations (CVO) Commercial Vehicle Operations applies ATMS, ATIS, and AVCSS technologies in the commercial vehicle operations such as trucks, towing trucks, taxies, paramedics, cranes, and other commercial fleets. The onboard GPS provide the real-time vehicle location information and wireless communication transmit the information back to control centers or dispatching centers. This location information is an important input to a fleet management system. After processing vehicle Introduction to ITS and NTCIP Figure 4. Dynamic bus information systems Figure 5. Onboard monitoring devices (Source: The Intelligent Transportation Systems of future) Figure 6. Driver monitoring systems (Source: The Intelligent Transportation Systems of future) 37 Introduction to ITS and NTCIP location information, the fleet management system can assign their commercial vehicles in a more efficient way to improve their productivities. Logistics companies, for example, see CVO as a critical system: freight being handled will be efficiently allocated in proper vehicles and be precisely tracked once they are en route. CVO enables them to reduce their costs while providing better delivery services (Figure 7). Related CVO technologies and applications include Automatic Vehicle Monitoring (AVM), Automatic Vehicle Locating (AVL), Weightin-Motion (WIM), Electronic Toll Collection (ETC), Automatic Vehicle Identification (AVI) and Automatic Cargo Identification (ACI), fleet dispatching systems, etc. Electronic Payment Systems (EPS) Electronic Payment Systems or Electronic Toll Collection (ETC) charges drivers/vehicles through the communications between onboard units (OBU) and roadside units. Conventionally tolls are collected by hands and that requires vehicles slowing down and preparing exact amount of fares otherwise incidents or delays will be incurred. On the contrary, EPS has the following features Figure 7. Commercial vehicle operations 38 to make itself a better solution such as vehicle no need to slow down, no incurred delays, lower long-term operating cost, no need to handle cash, and increased safety and facility utilization. See Figure 8. The other use of EPS is e-tickets for transit systems such as buses, taxies, parking, and airlines. If being combined with other payment systems successfully, users only need to carry one card to travel around without carrying a lot of cash or waiting in a manual operation line. Related EPS technologies include automatic vehicle identification (AVI), microwave/infrared Figure 8. EPS (ETC) in Taiwan Introduction to ITS and NTCIP sensing, Radio Frequency IDentification (RFID), electronic money, etc. See Figure 9. Emergency Management Systems (EMS) Emergency Management Systems is a system designed for emergencies. If an emergency such as a fire happened, EMS determines the dispatching and routing of fire trucks and ambulances, and these vehicles will be given priority to go to the scene. EMS utilizes the real-time traffic data from control centers, determines the best Figure 9. Automatic vehicle identification (Source: CECI, Taiwan, 2007) route with the minimum delays, and purposely adjusts the traffic signals along the route. Other agencies such as police will be notified as well to assist the rescue actions. EMS also determines the impact area and evacuates people within the area if necessary (Figures 10 and 11). Related EMS applications and technologies consist of Automatic Vehicle Location (AVL), Geographic Information Systems (GIS), automatic incident detection, Highway Advisory Radio (HAR), etc. Vulnerable Individual Protection Systems (VIPS) As shown in Figure 12, Vulnerable Individual Protection Systems (VIPS) protects the safety of handicaps, senior people, pedestrians, bikers, motorcyclists, etc. For example, dedicated traffic signals are designed for the blinds and handheld navigation devices on a bicycle or motorcycle that give warnings when a big truck is close from the behind. Related VIPS technologies include navigation systems, voice-warning traffic signals, etc. Figure 10. Automatic image detection system to identify incidents 39 Introduction to ITS and NTCIP Figure 11. Automatic incident detection-tunnel fire (Source: CECI, Taiwan, 2007) Figure 12. Vulnerable individual protection systems Information Management Systems (IMS) As shown in Figure 13, Information Management Systems is the backbone of the Intelligent Transportation Systems. IMS use all relevant information and communication technologies to collect, transmit, archive, process, analyze and share data. Data such as travel time, travel speed, vehicle volume, travel origin-destination, ridership, vehicle location, etc. are further analyzed to become usable information that will be delivered to users 40 and agencies and facilitate decision making. Users use this information to plan their trips to increase safety, efficiency, and reduce the travel time and/ or travel expenses. Agencies use this information to better manage their vehicles, terminals and to reduce the operating cost. Authorities use this information to design, modify or fine-tune their strategies, facilities to optimize the system performance. Without IMS, ITS designs, strategies, approaches and subsystems cannot be realized to the extent that they were designed to be. Introduction to ITS and NTCIP Figure 13. Traffic control center and IMS interface (Source: CECI, Taiwan, 2007) Future development of intelligent Transportation systems After few decades’ efforts, ITS provides a lot of benefits to users, agencies, authorities, and the environment. There are still a lot of issues to be solved and a lot of improvements to be made. Along with the development of new advanced technologies, ITS will provide us a better transportation environment that is also sustainable in the long run. nTcip NTCIP is different from the past practice of transportation management protocols in that it is not a single communications protocol designed for one purpose. Rather, it is a whole family of protocols covering from simple Point-to-Point protocols to complicate objects oriented techniques. This is due to several reasons: the diversity of the applications where NTCIP will be deployed, the resulting diversity of application specific characteristics, such as type and quantity of data to be transferred, the criticality of data transfer times, acceptable cost of communications infrastructure, and the criticality of data security and integrity issues. Interoperability and interchangeability are the key goals of the NTCIP open-standards effort. The term interoperability refers to the ability for multiple devices, often of different types, to work together as a single system for certain common purposes. For example, using the same communications channel to interconnect a management system with traffic signal controllers, dynamic message signs, video surveillance controls, and other devices. The terms interchangeability generally refers to the ability to use multiple brands of a device on the same communications channel, as well as the ability to swap them out. For example, the ability to put any brand of NTCIP-compatible traffic signal controller in the same system at the same time. 41 Introduction to ITS and NTCIP communications standards for c2F and c2c NTCIP provides two different types of ITS communications standards. The first type is between an ITS management center and multiple control or monitoring devices managed by that center. Below are some examples of this type of communications: • • • • A traffic signal management system communicating with signal controllers at intersections; A traffic management system controlling dynamic message signs, advisory radio transmitters, and environmental sensors on roadways; A transit management system communicating with passenger information signs on transit vehicles and at transit stations and stops; A highway management system communicating with detectors and ramp meters. This type of communication is called centerto-field, since most applications of this type involve a computer system at a management center communicating with various devices at the roadside or on vehicles. Protocols intended for these applications are often used in the environment where a central management station routinely polls each field device, as in the most common case of multiple field devices sharing a communications channel. The second type of communication involves messages sent between two or more ITS management centers. Below are some examples of this type of communications: • 42 Two or more traffic signal management centers exchanging information to achieve coordinated operation of traffic signals managed by the different centers and to enable personnel at one center to monitor • • • the status of signals operated from another center; A transit system reporting schedule adherence exceptions, to a transit customer information system or to a traveler information system, while also asking a traffic signal management system to instruct its signals to a behind-schedule transit vehicle; An emergency management system reporting an incident to a highway management system, to a traffic signal management system, and to a traveler information system; and A highway management system informing an emergency management system of a warning message in response to its notification of an incident. Such communication is called center-to-center, although two or more of the systems may be located within the same location or building. C2C involves peer-to-peer communications between any number of systems in an interconnected network which is similar to the Internet where any center can request information from, or provide information to, any other centers. It is also possible to use such protocols for communication to and between field devices, as well as between computer systems. Although both C2F and C2C communications can involve an operator making requests or issuing instructions, one of the features of the NTCIP protocols is their support for continuous and automatic functionality using pre-defined data transmissions. 3.2.1.1 Center-to-Field Protocols NTCIP provides three application level protocols for C2F communications: the Internet’s Simple Network Management Protocol (SNMP) (IEEE Std 828-1998, 1998; Feit et al., 1995; Stallings et al., 1996), the Simple Transportation Management Protocol (STMP) and the Simple Fixed Introduction to ITS and NTCIP Message Protocol (SFMP), which is currently under development. • • • Simple Network Management Protocol: SNMP provides a simple, but bandwidthinefficient protocol for C2F applications, based on the Internet protocol of the same name (SNMP). It is suitable only for networks with high bandwidth, or low volumes of messages. SNMP has been designed by the Internet community to run over UDP/ IP, but it can be forced to run over TCP/IP or T2/NULL. Simple Transportation Management Protocol: STMP was developed specifically for use in the transportation industry. It is an extension of SNMP that allows C2F messages to be sent more efficiently using dynamic objects. Stacks based on this protocol are suitable for networks with low bandwidth and high volumes of messages, including such traffic signal systems where a central computer is directly connected to field devices, without the need to route the information through some other devices such as an on-street master in a closed loop system. STMP has been designed to run over T2/Null since it supports low bandwidth links, but could also be used over UDP/IP or TCP/IP if there is sufficient bandwidth. Simple Fixed Message Protocol: There is a need for having a bandwidth efficient protocol for low-end field devices, like closed circuit camera controllers. NTCIP is currently developing SFMP to meet this need. These three protocols use the same get/set messaging paradigm as that used in SNMP. Although with the same base data elements, they differ in the level of complexity to implement and the types of services. Table 2 summarizes the comparisons among SNMP, STMP, and SFMP, while Figure 14 demonstrates the major advantages of these protocols. STMP is the most bandwidth efficient option currently available and includes full support of SNMP for infrequent messaging demands. It includes SNMP as a subset, so that any management system that implements STMP can also communicate with a device that supports SNMP. It also requires the use of SNMP to define dynamic objects. Infrequent messages requiring additional security can be sent using SNMP. STMP is the most flexible and bandwidth efficient option. The advantage of STMP is its support for dynamic objects which are combined with a more efficient encoding scheme, dramatically reduce the packet overhead relative to SNMP. Dynamic objects can also enable the user to define custom messages that are composed of any number of individual data elements. However, these data elements will have to be defined in both the center and the field devices in order to work properly. The field devices that use any particular subnetwork protocol can share the same communications with other devices using the same Table .2 SNMP, STMP and SFMP comparisons SNMP STMP SFMP Currently Under Development Ease of implementation Easy Hard Message set Support Limited to 13 Supports routing and dial-up Options Options Can send any base data element? Yes Yes Bandwidth efficiency-inverse of packet overhead Worst Best 43 Introduction to ITS and NTCIP Figure 14. C2F protocols protocol. It does not matter whether such devices are from different manufacturers or are totally different devices, e.g., a light signal and a dynamic message sign. Each device is assigned an address that is unique in that channel. The management system can communicate with any of the devices at any time by sending a message addressed to that device. However, when using Point to MultiPoint Protocol (PMPP), the management system can communicate with only one of the devices on the channel at a time. As a function of SNMP and STMP, devices can only send a message to the management system when requested by the management system. The NTCIP protocols enable broadcast messages for all devices, and no devices can reply to a broadcast message. When planning a C2F communications network using NTCIP that involves continuous polling of field devices, e.g., a traffic signal system or transit fleet system, it is important to consider the relationship between the following key variables: • • • • • • • • 44 Transmission schemes. Time between devices or between polling cycles. Transmission rate. Frequency of communication. Transmission delay. Response delay in the field device. Frequency of each type of message. Length of message to be sent. • Number of devices sharing the same line or channel. The first seven of these variables will determine the total time needed to communicate with each device. Although STMP is designed for use with communications channels that use a slow transmission rate, it is not as bandwidth-efficient as most proprietary protocols used in the past. With existing communications infrastructure, it may not be possible to maintain the same polling period with the same number of devices per channel. This is due to the fact that proprietary protocols are optimized for each manufacturer’s equipment and consist of very few fixed short messages without any flexibility in terms of changing these messages, while standard protocols are flexibly designed to accommodate all needs and a wide variety of information and messages in a multi-manufacturer environment. However, careful design can usually find a reasonable compromise between the principal variables. Higher available bandwidth, bit rate, yields fewer compromises or required trade-offs. If new communications infrastructure can be provided, it will allow for additional channels and/or higher transmission rates. Center-to-Center Protocols NTCIP originally provided two application level protocols for C2C communications, DATEX (Data Exchange) and CORBA (Common Object Requesting Broker Architecture), which were found necessary to meet the variety of requirements for inter-system data exchanges. Recently, there has been increased interest in using XML and Web Service technologies for C2C links due to its simplicity and the wide accessibility of tools to provide these services. However, it is important to determine where to deploy these protocols in the network, with some centers acting as bridges, or translators, among different protocols. DATEX was designed to provide simple and cost-effective solutions for the following needs: Introduction to ITS and NTCIP • • • • Systems requiring real: time, fast data transfer; Systems with limited communications bandwidth but high data transfer load; Systems with infrequent event driven exchanges over dial-up links; and Non-object oriented systems. DATEX provides a general-purpose C2C data exchange protocol stack. It uses pre-defined messages transmitted by the base Internet protocols (TCP/IP and UDP/IP) in a peer-to-peer network. The base standard at the application level is an ISO standard. On the other hand, CORBA provides several features to support networks connecting object oriented systems, and assuming sufficient processing power and communications bandwidth are provided. Object oriented software can take full advantage of CORBA and implement it easily; this is much more difficult to achieve with traditional procedural software. CORBA is a general purpose C2C communications protocol based on the computing industry standard of the same name. For object-oriented systems, it enables a higher degree of integration and some services not provided by DATEX, but it may not be suitable for near real-time applications and loosely coupled systems. The wide availability of XML tools and large market have generated the market interest in XML. It is especially suited for systems requiring limited, simple data exchanges over communications links with sufficient bandwidth and processors with sufficient processing time available. However, there are no current transportation industry standards for the use of XML. The NTCIP effort continues to monitor the maturity of XML in an effort to determine its suitability for future use in the transportation industry. C2C networks allow each system to request available information from any other systems. Each system can be configured to either accept or reject certain requests. The data sent can either be informational or can constitute a command to take some actions. Consider a message sent from one traffic signal system to another and containing a signal timing pattern number. In DATEX, depending on the message type, it could represent a command to implement that timing pattern at a particular traffic signal or group of signals, or it could represent a status report indicating that this timing pattern was just implemented at a particular traffic signal or group of signals. The user can also establish standing subscriptions for data. In DATEX, these subscriptions can specify that data be sent one-time-only, periodically, or repeatedly on occurrence of some event as defined in the subscription. Each subscription message has a corresponding publication message. Unless the subscription is a one-time request, the data will continue to be automatically published repeatedly until the subscription is cancelled, or until a predefined end date specified in the subscription. A system can use CORBA to automatically discover data availability and shared control options available from other systems. These other systems use the CORBA framework to publish their capabilities and services offered, accept registration requests from authorized clients, and then deliver those capabilities and services to those clients on demand. For example, a CORBA traffic management system that owns a CCTV can offer to provide: (1) the images acquired as (a) snapshot, or (b) streaming video, and/or (2) allow remote control movement of that CCTV. The system owning the CCTV is the server and the system asking for the images, and/or control of the CCTV is the client. This example also serves to illustrate a typical use of a subscription such as “send me a new snapshot image from CCTV every minute” stated in the proper terms for that CORBA system- assuming the requester is authorized that service, the expected result is fairly obvious. C2C communications require a peer-to-peer network connection such as a local area network 45 Introduction to ITS and NTCIP or a wide area network between the involved centers. Local area networks typically use agencyowned twisted pair cable or fiber optic cable. Wide area networks typically use commercial telecommunications links such as frame-relay, fractional T1 leased lines, packet radio, leased “virtual private networks”, ISDN, or similar modems over “plain-old telephone” lines. Any type of communication link can be used, as long as it enables use of the Internet transport and routing protocols (TCP/IP and UDP/IP) and has sufficient bandwidth for the planned communications load to achieve the desired operational performance (this is based upon frequency, size of messages to be exchanged, and latency issues encountered when using C2C systems). For DATEX and CORBA, the base protocols have been defined, that is, how to exchange data, but the standards defining the data to be exchanged have not reached a state of maturity. The XML approach is even less mature in that the industry has not agreed on the exact rules on how to exchange the XML documents. Any recent Figure 15. NTCIP and the ITS architecture 46 deployment should consider the impacts that this may have on the long-term maintainability of a system. The best solution is still likely to deploy one of the recognized standards, but the agency should realize that a future project would likely be required to upgrade the software to address any included features affected by revisions in order to achieve the final mature standard. The role of nTcip in the iTs architecture NTCIP defines a family of general-purpose communications protocols and transportation specific data dictionaries/message sets (IEEE Std 14881999, 2000; IEEE Std 1489- 1999, 1999; IEEE Std 1512-2000, 2000) that support most types of computer systems and field devices used in transportation management. Applications for NTCIP are generally divided into two categories: C2F and C2C. The former, normally involves devices at the roadside, communicating with management software on a central computer. Introduction to ITS and NTCIP C2C applications usually involve computer-tocomputer communications where the computers can be in the same room, in management centers operated by adjacent agencies, or across the country. The role of NTCIP in the National ITS Architecture is illustrated in Figure 15. For both C2F and C2C applications, NTCIP supports systems and devices used in traffic, transit, emergency management, traveler information and planning/data archiving systems. Figure 16 illustrates how various transportation management systems and devices can be integrated using NTCIP. The following are examples of systems and devices that can take advantage of NTCIP: • Center-to-Field ◦ Traffic signals ◦ Dynamic message signs ◦ On-board sensors and controllers ◦ Environmental sensors ◦ Ramp meters • ◦ Vehicle detectors ◦ Highway lighting control Center-to-Center ◦ Traffic management ◦ Transit management ◦ Incident management ◦ Emergency management ◦ Parking management ◦ Traveler information ◦ Commercial vehicle operations regulation Many applications of NTCIP are related to real-time communications and involve continuous transmissions of data or commands while historical data can be sent using NTCIP, other communication standards, especially e-mail and file transfer protocols developed for the Internet, may also be used. Human-to-human communications are generally better served by fax/telephone and Internet protocols, but basic support is also provided in the NTCIP C2C protocols. Figure 16. Example of ITS integration using NTCIP 47 Introduction to ITS and NTCIP The nTcip Framework When options are available with layered protocols, the options can be diagramed in a “framework.” Figure 17 illustrates the framework for NTCIP. The diagram shows the different protocols that can be chosen at each level and which ones are compatible. However, not all compatible configurations make sense, and there are mutually exclusive choices. A particular message transmission can use at least one protocol from each level of the NTCIP framework. The series of protocols used in the message transmission is called a “protocol stack.” It is possible for a pair of electronic devices to exchange some messages using one stack and other messages using a different stack, though usually, such stacks will differ only at one or two levels or sublevels. As shown in Figure 17, the lines connecting standards at different levels show optional standards at each level. If there is a continuous line from one standard to another, then they are compatible and can be used together Figure 17. NTCIP standards frameworks 48 as part of a protocol stack. There is also another way to classify the NTCIP standards: primary, supporting, and base standards protocols. • • A primary standard applies directly and specifically to the device or component subsystem being implemented. For example, standards 1203 and 1204 are primary NTCIP standards and the Standard 1203 applies specifically to DMS, and 1204 to ESS. A supporting standard applies in general to more than one device or component subsystem implementation. For example, the NTCIP Standard 1201 Global Objects standard applies to all devices and component subsystem implementations that use or require features such as identification and location of equipment, global time, and event detection or scheduling. Thus, standard like Standard 1201 is a supporting standard. Both primary and supporting Introduction to ITS and NTCIP • standards typically apply at the C2F or C2C information layer (Figure 17). A base standard and protocol applies to the Application, Transport and Sub-network levels. These standards define NTCIP unique capabilities for protocol and data transport choices to complete the design of an operational deployment. These standards is different from both primary and supporting standards, since the data being exchanged is irrelevant. These standards are unaware and largely unaffected by their use in a signal control, DMS, and ESS applications. When deploying an NTCIP-based system, protocols have to be chosen. Figure 18 illustrates an example if C2F protocols stack choice that can be defined using NTCIP standards. A stack is a subset of the overall NTCIP framework-a selected route through the levels, given the choices available. Some stacks include two standards at some levels, which usually mean the protocol can use either of the optional standards. NTCIP protocols generally offer further options within most of the standards. Examples of sub-options within a standard are: which subset of messages are supported, or which bit rate is used at the physical interface. nTcip communication levels NTCIP uses a layered approach to communications standards, similar to the layering approach adopted by the Internet and the International Organization of Standards (ISO). In general, data communications between two computer systems or other electronic devices can be considered to involve the following primary layers, called “levels” in NTCIP, to distinguish them from those defined by ISO and the Internet. The NTCIP standards publication numbers are grouped in number ranges to indicate the standard type and the level where the standard goes. • Information level: This level contains standards for the data elements, objects and messages to be transmitted, e.g., TCIP, Figure 18. Example center-to-field stack 49 Introduction to ITS and NTCIP • • • • NTCIP 1200 series Standards Publications, MS/ETMCC. Application level: This level contains standards for the data packet structure and session management, e.g., SNMP, STMP, DATEX-ASN, CORBA, FTP. Transport level: This level contains standards for data flow control, packet reassembly and routing when needed, e.g., TCP, UDP, IP. Subnetwork level: This level contains standards for the physical interface, e.g., modem, CSU/DSU, and the data frame encapsulation method, e.g., HDLC, PPP, Ethernet, ATM. Plant level: This level consists of the physical transmission media used for communications, e.g., twisted pair copper wire, coaxial cable, fiber optic cable, wireless. It should be noted that the plant level is an infrastructure choice and not a standards selection choice. However, the plant level selection will have an impact on the subnetwork level selection to which it must interface. Information level standards used in ITS are unique to the transportation industry. The National ITS Architecture and much of the standards development effort for ITS involve identification of required data elements and the definition of their use for all the different domains and functions within ITS, e.g., traffic, transit, traveler information, emergency management. At the Application, Transport and Subnetwork levels, ITS can frequently use existing standards used by the broader computer and telecommunications industries. Below the Information level, the NTCIP standards deal with choosing which existing standards are to be used in ITS. The Internet standards have been adopted where possible. The NTCIP standards specify which options to use where alternatives are available in some standards. NTCIP has not had to develop significantly new standards in 50 these areas. The two major exceptions are the protocols that support: • • • • • Slow speed, high frequency communications links as found in 1200 bps, once-persecond traffic signal systems. A simplified Publish-Subscribe C2C protocol. NTCIP has extended existing standards or developed entirely new protocols as needed in cases where ITS has special protocol requirements. The two areas include: Continuous, automated, real-time exchange of large volumes of small data packets in a many-to-many multi-agency network. Continuous high volumes of real: time data sent to and from embedded processors in roadside or on-vehicle equipment sharing the same, low-speed, data channel and requiring low latency. Through a layered combination of existing communications standards and a few new standards developed specifically for ITS, NTCIP provides a family of communications protocols that serve many of the common needs in ITS transportation management. The levels shown in the framework are somewhat different from communication stack layers defined by the ISO’s Open Systems Interconnect seven-layer reference model and other standards developing organizations. The NTCIP stack extends beyond the communications stack to include informational data and interfaces to the actual communications infrastructure. The levels and terminology used in NTCIP were chosen for simplicity and ease of understanding by readers, and related to typical applications in the transportation industry. With the many diverse requirements of NTCIP, it is not surprising that we looked at the ISO OSI Reference model to help us define the framework for the new family of standards. Although OSI communications protocols are not widely used, the Introduction to ITS and NTCIP Figure 19. Level mapping of OSI layer to NTCIP level mapping layered model remains. The OSI model breaks the communications process into seven well-defined layers. Each layer has a defined purpose, generally independent of adjacent layers. Figure 19 shows how the NTCIP Information, Application, Transport, Subnetwork and Plant Levels loosely relate to the OSI model and are described as follows: • • NTCIP information level: This level defines the meaning of data and messages and generally deal with ITS information. This is similar to defining a dictionary and phrase list within a language. These standards are above the traditional OSI sevenlayer model. Information level statndards represent the functionality of the system to be implemented. NTCIP application level: This level defines the rules and procedures for exchanging information data. The rules may include definitions of proper grammar and syntax of a single statement, as well as the sequence of allowed statements. This is similar to combining words and phrases to form a sentence, or a complete thought, and • • defining the rules for greeting each other and exchanging information. These standards are equal to the Session, Presentation and Application Layers of the OSI model. NTCIP transport level: This level defines the rules and procedures for exchanging the Application data between source and destination on a network, including any necessary routing, message disassembly/ re-assembly and network management functions (Buede et al., 2000). This is similar to the rules and procedures used by the telephone company to connect two remotely located telephones. Transportation level standards are roughly equivalent to the Transport and Network Layers of the OSI model. NTCIP subnetwork level: This level defines the rules and procedures for exchanging data between two devices over some communications media. This is equivalent to the rules used by the telephone company to exchange data over a cellular link versus the rules used to exchange data over a twisted pair copper wire. These standards 51 Introduction to ITS and NTCIP • are roughly equivalent to the Data Link and Physical Layers of the OSI model. NTCIP plant level: This level is used to provide a point of reference to those learning about NTCIP. The Plant level includes the communications infrastructure over which NTCIP communications standards are to be used and will have a direct impact on the selection of an appropriate Subnetwork Level for use over the selected communications infrastructure. The NTCIP standards do not prescribe any one media type over another. To ensure a working system, deployers must specify and/or select an NTCIP protocol or profile at each level. Most of the standards in the lower levels are existing commercially available standards used in the telecommunications industry and were not developed uniquely by NTCIP, although NTCIP often specifies which sub-options within those standards are to be used. The majority of standards unique to Intelligent Transportation Systems are found in the Information Level and Application Level shown at the top of Figure 17 and Figure 18. Each NTCIP protocol stack involves a mixture of standards, with at least one from each level. major protocols in nTcip protocol stacks The first NTCIP standards developed were those intended for C2F applications. This involved a new application level standard called Simple Transportation Management Protocol, a new transport level standard called the Transportation Transport Profile (T2 or T2/NULL), and several sets of new standard data elements (Michael et al., 2004; IEEE Std 1489-1999, 1999) called “object definitions” at the information level. The initial NTCIP C2F protocol development also involved references to three existing standards: 52 • • • Point-to-Point Protocol (PPP) A customization of the High-level Data Link Control (HDLC) standard at the subnetwork level, known as the Point-toMultiPoint Protocol and SNMP standard at the application level. Each standard specifies one or more protocols to be used at a given level, and the suboptions allowed or required within each of those standards. A standard publication will typically reference one or more “normative standard” publications-other publications that contain additional relevant specifications for the standard(s). A referred normative standard may be another NTCIP publication, if the standard was developed by NTCIP, or a publication developed by any other standards development organization. For a particular application of NTCIP, the user must select, in the procurement specifications, which elements are desired at each level. e.g., select from the options called out in one or more profile publications for each level. The set of selections and options for base standards and protocols for all levels is referred to here as a “protocol stack.” Each NTCIP protocol stack will have different characteristics, and a stack that works well for one application or communications environment may not suit another. Application level standards that NTCIP has defined are briefly described below. As pointed out in Fig. 3.17 and 3.18, these application profiles can be combined with certain transport profiles. Two electronic devices will be better able to communicate interchangeably with each other, if they use the same communications protocol stack, the same information level data dictionaries and message sets, and implement the same desired options defined in each of these selected primary, supporting, and base standards and protocols. In addition to specifying a protocol stack, the system designer must also choose between various options and alternatives available in the selected stack. These options exist in both C2C and C2F protocol stacks. Major options, such as Introduction to ITS and NTCIP which protocols to support at each level in the communications stack, are sometimes grouped according to conformance levels, while others are individually selectable. Most manufacturers and system suppliers typically offer features that go beyond the standard. To make use of such features, it is necessary to specify the inclusion of manufacturer-specific data elements or messages as extensions of the standards when procuring a management system. The decision by an agency to use features above and beyond the standard should be taken only with the understanding of the potential impacts. These impacts could be considerable in the long term. These options may, in effect, result in the purchase of proprietary systems. Part of the decision must include how many of these features that will be allowed. migration from legacy systems to nTcip Since interoperability and interchangeability are two key goads of NTCIP, the inability to update older equipment should never stop an agency from replacement or migration strategies to make full use the benefits of NTCIP conformant implementations. For example, a central system whose current field devices cannot be updated could be expanded to run NTCIP protocols on some communications channels while the older equipment is maintained on others. There is a model for a three-step migration from legacy systems to NTCIP. Initially, the proprietary interface details may or may not be known. Then, there is some intermediate state and some period of time where the operational system consists of a mixture of the legacy systems and the newer NTCIP hardware. There may be a common communications channel for legacy and NTCIP devices. The central control system may be separate or combined; it may run on the same computer or on separate computers: this is determined by the scope of the project to accom- plish these migration steps. Pursuit of a migration strategy towards the use of open standards starts to minimize the use of proprietary communications and begins to maximize the use of NTCIP. Finally, at some future point, the migration is completed and NTCIP is fully deployed, having replaced all now retired legacy systems. NTCIP and non-NTCIP devices may be mixed on the same channel. Thus, all devices sharing a channel must be upgraded simultaneously. A center that communicates with both NTCIP and non- NTCIP devices will need to use a different communications port for NTCIP devices and for non-NTCIP devices, and will need to support both protocols. Therefore, the mixed devices listening on the shared communications channels must recognize and react only to those data elements and commands intended for them individually, and must also not produce unpredictable results in response to any other data traffic on the channel. For example, the most likely and simplest solution in the traditional closed-loop traffic signal systems is to limit each field master to one protocol. Only field masters with NTCIP-compatible controllers would be upgraded to support NTCIP. This avoids the need for field masters to simultaneously support two protocols on two separate ports. The center could communicate with field masters using a different protocol than that used by the field master to communicate with controllers. As with the controllers and the field master, the center’s software will need to be modified to add support for an NTCIP protocol, if NTCIP is to be used for communications with field masters. Any upgrade of an existing system to add support for NTCIP should be designed in consultation with the system provider. Each provider should adopt an upgrade or migration strategy that is most efficient for the majority of its customers. If a customer wants a unique arrangement, that customer may have to pay the full cost of the software modifications, whereas the cost of the general solution can be spread among many customers. One approach to the introduction of 53 Introduction to ITS and NTCIP NTCIP in a C2F system is to operate NTCIP and non-NTCIP systems during a transition period. Field devices can gradually be switched over from one to the other as they are replaced or their software is upgraded. If the current system is quite old and upgrading it for NTCIP is not practical, this transition should be done as part of a general system upgrade. nTcip implementation examples As shown in Figure 20, NTCIP have been deployed across U.S.A in several states. The initial deployment of NTCIP-conformant equipments was conducted by the Virginia Department of Transportation (VDOT), USA in a case study of their Variable Message Sign implementation in NTCIP-9002 Version 01.04, in September, 1999. AASHTO, FHWA, ITE and NEMA are currently sponsoring this case study update. This effort, presented as a case study amendment, focuses on insights gained over the three years of deployment since the initial case study was performed. Figure 20. NTCIP projects in U.S.A. 54 Specifically, this amendment will address Agency issues concerning current implementation efforts and their needs based upon experience gained through NTCIP deployment experience. The Washington State Department of Transportation (WSDOT), USA implemented a Variable Message Sign (VMS) system in 1999. The formal name of this project is the “NTCIP VMS Software Upgrade.” The purpose of the project was to modify the existing traffic management system to support selected protocols from the NTCIP protocols and to purchase two new NTCIP compliant variable message signs. The vendor was American Electronic Sign Co (AES). Additionally, the AGENCY hired a programming contractor (PROGRAMMER) to enhance the central system software. The City of Phoenix, Arizona, USA initiated a project to enhance their traffic signal system as shown in Figure 20. This project included two distinct parts, the replacement of the central traffic control system, and the upgrade and purchase of additional traffic signal controllers. The NTCIP Introduction to ITS and NTCIP was specified as the communications protocol of choice for both. The central system, called Phoenix ATMS was also required to control additional field devices, however, without the requirement to utilize the NTCIP communications protocol. The City of Lakewood, Colorado, USA initiated a project to enhance its traffic signal system. This project included two distinct parts, the replacement of the central traffic control system, and the upgrade and purchase of additional traffic signal controllers. The NTCIP was specified as the communications protocol of choice for both. The central system, called Lakewood ATMS was also required to be extensible in order to add future capabilities for controlling field devices such as NTCIP Dynamic Message Signs and non-NTCIP closed circuit television cameras. The City of Mesa, AZ initiated a Request for Proposal to upgrade their existing SONEX system, which included a requirement for NTCIP, i.e., “Support NTCIP for communication to TS-2 controllers”. However, the proposed cost was so much higher than estimated, that the Agency withdrew the RFP. In 1997, the Agency again issued a Request for Proposal, this time requesting to replace the entire signal system including the central system and the signal controllers. A phased approach was to be used running temporary parallel central systems. The NTCIP was specified as the communications protocol of choice for the new components. The central system, called Mesa ATMS was also required to control additional field devices and provide an interface to the SYNCHRO software. In England, a project, called UTMC (Urban Traffic Management and Control) had been implemented and evaluated. However, the NTCIP definition cannot support the England metropolis area at present the material transmission demand. If NTCIP is applied to the European area, the MIB file must be revised or established to conform to the application demands of the England area. UTMC indentified that SNMP communications produce serious overhead problem. Instead of using SNMP, the STMP communication is suggested to be used in England. reFerences Buede, D. M. (2000). The engineering and design of systems: Models and methods. New York: Wiley, Inc. Feit, S. (1993). TCP/IP: Architecture, protocols and implementation. New York: McGraw Hill, Inc. Feit, S. (1995). SNMP: A guide to network management. New York: McGraw Hill, Inc. Institute of Transportation Engineers Management and Operations of Intelligent Transportation Systems. (2003). ITS standards overview. Internet Librarian, D. I. S. A. (1997). US-DOD Internet related standardized profiles. Retrieved from http://www-library .itsi.disa.mil/ Michael, A. (2004). Guide to the IEEE 1512 family of standards. Washington, DC: Institute of Electrical & Electronics Engineer. Research and Innovative Technology Administration (RITA) & U.S. Department of Transportation. (US DOT). (2002). National ITS architecture. Retrieved from http://itsarch.iteris.com/itsarch/ Stallings, W. (1996). SNMP, SNMPv2 and RMON. Reading, MA: Addison-Wesley Publishing Company, Inc.f NTCIP Standard 9001. (2002). The NTCIP guide. ITS Standards Outreach, Education and Training Program, Institute of Transportation Engineers. (2006). Center to center communications. IEEE Std 828-1998. (1998). IEEE standard for software configuration management plans. 55 Introduction to ITS and NTCIP IEEE Std 1489-1999. (1999). IEEE standard for data dictionaries for intelligent transportation systems. IEEE Std 1488-1999. (2000). IEEE standard for message set template for intelligent transportation systems. 56 IEEE Std 1512-2000. (2000). IEEE common incident management message sets for use by emergency management centers. Section 2 Embedded System Architecture and Communication Protocols 58 Chapter 4 Vehicular Embedded System Architecture Chung-Ping Young National Cheng Kung University, Taiwan, R.O.C. absTracT The dramatic advancement of IC technologies makes electronic devices be smaller and run faster, so they are able to implement more functions in a limited space. The car electronics play an increasingly important role in automobile industry, and the embedded system has already been extensively employed for improving the operation and performance of vehicles, such as safety, comfort, convenience, and energy consumption. In terms of electronic system, an automobile is a distributed embedded system, and the control messages to each electronic control unit (ECU), go through in-vehicle networks. An ECU is a computing system, integrated with a data acquisition module or an electromechanical driver. A variety of ECUs implement versatile functions, such as powertrain, antilock braking system (ABS), traction control system (TCS), adaptive cruise control (ACC), and electronic stability program (ESP), etc. Sensors provide measurements of specific vehicle parameters in a format suitable for the digital microcontroller, while actuators are electrically operated devices that drive electromechanical components. Human machine interface is the input and output of vehicle operations to users. inTroducTion Transportation of humans and objects have been playing an important role in our daily lives since civilization first formed and needed new means of reaching destinations. The invention of efficient transportation greatly reduced the time and labor once required and in addition largely extended the living environment that people can reach. The more time and labor for transportation is saved, the more leisure time people will have. Animal-power or natural resources have been the driving force of transportation for a long time. After the steam engine was invented, the automobile started a new era. The mass production of the Ford model T cre- DOI: 10.4018/978-1-60566-840-6.ch004 Copyright © 2010, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global is prohibited. Vehicular Embedded System Architecture ated the modern automobile industry and made the automobile more affordable. The basic structure of the automobile has not changed much, but evolving technologies has kept improving its functions and performance. The construction of traffic networks and mass production of automobiles have made the automobile the most important land based transportation carrier. The usage of automobiles is usually associated with the growth of economy and industry of a nation, so the population ratio that owns automobiles in a developed country is larger than that in a developing country. When the economy grows, vehicle as a transportation tool becomes more affordable and popular, for instance China or India. When people use automobiles in their daily lives, they demand not only mobility, but also safety, comfort and convenience. These are some design factors that manufacturers have to put into aspect when enhancing functions by introducing and developing new technologies. For a government to provide a modern transportation system, it has to build not only a traffic network, but also an infrastructure to access more information to allow drivers and passengers to drive safer, more comfortably and with better convenience on the road. This is the vision of an intelligent transportation system. To achieve this goal, both the infrastructure and vehicles have to be facilitated with a modern electronic and information system. The evolution of an automobile shows more signs of adopting electronic devices. To enhance the features or performance, some mechanical components are replaced by wires and electronic devices, or simple electronic devices are enhanced by complex electronic control systems. An automobile consists of several control systems: power train, chassis, safety, body and information. Each system may have several subsystems distributed in different location of a car and are linked through in-vehicle networks. Since most modern electronic devices digitize signals and process them by software, embedded systems are employed for data processing and control in an automobile. The dramatic advancement of IC technology, which is described by Moore’s law, makes chips smaller, faster, and able to implement more functions. In terms of electronic systems, an automobile is a distributed embedded system, and the control messages to each distributed device go through the in-vehicle network. X-by-wire is becoming a new technical trend. From a top-down viewpoint, the ultimate goal of transportation is to develop an intelligent transportation system. The basic mobile unit is a vehicle, which is interconnected to other vehicles or backend service providers through vehicle-tovehicle or vehicle-to-infrastructure communications. The scope of this chapter is limited to the distributed embedded systems in a vehicle. A vehicle consists of a variety of electronic control units interconnected through an in-vehicle network, while each unit is an embedded system involving processor and memory along with other optional sensors, actuators, storage devices or human machine interface. saFeTy, comForT and conVenience Vehicles were developed for transportation. When vehicles become mandatory transportation tools in daily life, safety is the first issue. Road safety is related to the loss of human life and property and can be categorized into three areas: human, environment, and vehicle. Human and environment factors are out of the scope, and we will focus only on the vehicle. However, the enhancement for vehicle safety can sometimes compensate the inappropriate operation caused by human or environmental factors. Vehicle safety can be further separated into active safety and passive safety (Robert Bosch, 2006). The active safety mechanism is to prevent the happening of potential accidents, while passive safety is to 59 Vehicular Embedded System Architecture help drivers and passengers lower injury or death rates by accidents. Figure 1 shows the different categories of road safety, and goals for active safety and passive safety. Active safety systems, including antilock braking system (ABS), traction control system, and electronic stability program, enhance the stability and steerability of driving so that corrections or reductions toward inappropriate operations caused by humans or the environment can be made, therefore improving road safety. Adaptive cruise control not only provides the safety function of maintaining safe headway with the preceding vehicle, but also relieves the drivers need to frequently check the speed. A car accident happens through three phases: pre-crash, in-crash and post-crash stages. To prevent the happening of a tragedy, the safety mechanism must successfully work before the pre-crash stage. The active safety provides safe and convenient drivability to reduce accidents, and the passive safety provides pre-triggering to the protection device for the impact and postcrash communications. Figure 2 presents three Figure 1. Road safety categories 60 crash stages and the active/passive safety related safety mechanism. The lifetime of market penetration of one automotive system usually begins with an innovative technology applied to a high-end model vehicle. Later on this technology is recognized as a major improvement to environment or safety and required by law, or recognized as a mandatory feature by the customer. This product then becomes the standard equipment in all car classes. Seat belts, airbags, and tire pressure detection are regulated by legislation to be equipped on a vehicle, while blind spot detection and adaptive front lighting are trying to be regulated in Europe. The comfort and convenience are not as critical as safety for evaluating a vehicle. To provide a joyful and easy driving environment, there are important design factors to be considered when increasing the quality of driving and riding. Sometimes, some features of comfort and convenience are also correlated with safety improvements. For example, blind-spot warning produces a visual/ audible signal to notify the driver of potential dangers: a car approaching in the blind-spot area. Vehicular Embedded System Architecture Figure 2. Active and passive safety in different crash stages The driver can concentrate on the front view of the forward direction, but not be distracted by the vehicle on either side. neTworked embedded sysTems overview To fulfill a variety of functions in a vehicle, each electronic control unit (ECU) independently implements its designated function. The automotive system is constructed as a server-client architecture. ECUs can be treated as a peripheral device of the central console, and the digital data are communicated via the in-vehicle network. The central console sends the command to ECU and the ECU sends the status back to central console. A vehicle can be categorized by the following systems on vehicular network: powertrain, chassis, safety, body, central console, and infotainment. Figure 3 shows that all vehicle electronic control modules are networked through in-vehicle network. powertrain The powertrain system consists of engine control, automatic transmissions, traction control system (TCS) and an adaptive cruise control (ACC). The engine control is to gain higher engine power at a lower system cost level, while also offering improved fuel efficiency and reduced exhaust Figure 3. Networked vehicle electronic control modules 61 Vehicular Embedded System Architecture emissions. Transmissions convert the torque generated by the engine and the engine speed corresponding to the tractive force requirement. The automatic transmissions automatically take on control of starting, selecting the gear ratios and switching gears. The TCS prevents the wheels from spinning by reducing the drive torque at each driven wheel. The ACC system maintains the constant speed set by the driver, and will reduce the speed to follow a slower vehicle. It speeds up by electronically accelerating up to the setting via the engine management system or decelerates by electronically activating the brake system. chassis Chassis systems enhance vehicle stability and steerability. Electronic stability program (ESP) improves vehicle handling and braking response through programmed intervention in the braking system and/or powertrain. It integrates both ABS, preventing from wheel locking when brakes are applied; and TCS, preventing from wheel spinning when acceleration is applied. The suspension adjustments compensate the rise and fall of wheels on cornering or road irregularities to optimize vehicle stability. safety When the automotive system senses an impact, the passive safety systems activate. Front or side airbags are triggered and fully inflated in about 40 ms and vehicle occupant detection system drives the seat belt pretensioner to function in restraining the occupants in their seats. The ABS prevents the wheels from locking up when the brakes are applied by lowering the wheel brake pressures. body and comfort control Body and comfort control systems provide occupants convenience and comfort features, such 62 as power windows, power sliding/rear doors, door lock, wiper, ventilation, heater, air-conditioning, room lamp, and adaptive front lighting. central console Central console is the user interface to the automotive system. It involves dashboard projection of the vehicle status and button/knob to control the operation and setting. A concept central console will have only a graphical LCD display with touch screen panel integrated. The communication between the central console and other ECUs relies on the in-vehicle network, such as CAN/ LIN, FlexRay, and MOST (Paret, 2007). infotainment Infotainment system provides the contents of information and entertainment to occupants. Entertainment involves audio, video, radio, digital TV, or gaming. The infotainment contents can be retrieved from local mass storage devices or via radio, mobile communication, or wireless network to remote service/content providers. The navigation system provides location awareness and route planning functions by utilizing multiple xGPS systems. X-by-wire Since each ECU is located separately and connected through in-vehicle networking, the control of an electromechanical device is not by the conventional mechanical parts, but by the digital control message on the bus. X-by-wire implies that the operation on a vehicle is implemented through electrical wires, where x could be any operation that you name. This reduces the cost and space for mechanical parts, and the automotive system appears as a mobile vehicular network. Some examples of x-by-wire are steer-by-wire and brake-by-wire (Paret, 2007). Vehicular Embedded System Architecture elecTronic conTrol uniT overview During the technical evolution of automobile development, the relatively straightforward electromechanical operations in a vehicle cannot fulfill the complex safety or comfort requirements in a modern vehicle. The electronic devices have become more popular in the 1960s, since the dawn of the semiconductor era. The computing systems brought signal processing to a brand new digital world and then enhanced the system control. Because of the advancement of semiconductor technology, the market pushed the automotive systems to a more precise and complicated control in fuel injection, braking, and traction control. Electronic Control Unit (ECU) is basically a processor or computing system, integrated with a data acquisition device or an electromechanical driver. Sensors provide measurements of specific vehicle parameters in a format suitable for the digital microcontroller. Similarly, actuators are electrically operated devices that regulate signals to the device that directly control its output. The ECU is modeled as a closed-loop control system. The ECU performs a looped control by sensing the automotive system parameters, processing the computation for determination of appropriate reaction, and then outputting the control signal to adjust the operation of the automotive system. This control loop starts again, when the automotive system parameters change (Robert Bosch, 2006). Figure 4 demonstrates the closed-loop control of automobile steering under variable human, vehicle, or environmental parameters. Some examples of ECU implementations are described in the next few subsections. Due to the strict operation environment in a vehicle, some factors are carefully considered before developing an ECU. • Size: Though the dimensions of a vehicle seem to be large, it is designed to have most of the space reserved for driver and passengers. The physical size of an ECU has to be small and is integrated into the electromechanical module. The advancement of the semiconductor manufacturing process Figure 4. Closed-loop control of automobile steering under variable human, vehicle and environment parameters 63 Vehicular Embedded System Architecture • • • makes this miniaturization requirement feasible. Cost: Cost is always an important factor in promoting a product on the market. The electronic components are utilized more than ever in a modern vehicle where several ECUs can be utilized in an automotive system, so the cost of the electronic devices is also a concern. All in all, processor technology, IC technology, or component specifications determine the cost. Function: Due to different requirements for individual subsystems in an automotive system, ECUs are developed for various automotive functions. A dedicated computing system embedded in an ECU can optimize its control performance. Though the system architecture is similar, the hardware of sensors, actuator components, processor technology, and firmware of data acquisition, control and signal processing are different. Performance: Powertrain or brake controls are time-critical when driving on the road. A real-time operation, which guarantees the designated tasks can be finished before the critical deadline, is required. Vehicle speed is very high on highway and human reaction time is short, so the system • • performance is required to be efficient to ensure the appropriate control. Stability: The implementation of an ECU does not only require the performance but also the accuracy. This is to ensure the appropriate functioning of the system. The conventional mechanical control can be investigated by vision or instrumentation, but the operation of an electronic system with hardware and software is invisible and has to be tested through a suite of carefully defined test cases. However, this still does not guarantee the perfect functions without any errors. The fault tolerant or self-recovery mechanism must be considered to be implemented in the system to enhance the stability. Robustness: Due to severe operation environments in a vehicle, if the hardware is not robust enough it may cause the software to malfunction. The electronic components and the board design have to be developed on a stricter industrial standard for tolerating temperature, humidity, electrical noise, or electromagnetic interference. Figure 5 depicts the block diagram of an electronic control unit, while the blocks surrounding the computing platform are optional depending on the function of the ECU. Figure 5. A complete block diagram of electronic control unit 64 Vehicular Embedded System Architecture powertrain Powertrain management controls the engine and transmission. From the sensor inputs, the environmental information, driving-conditions, vehicle and user parameters are obtained for the powertain controller to determine the operation of the engine and transmission; then the powertrain controller commands the engine, traction controller, and transmission to generate the appropriate torque, slip, or gear ratio (Robert Bosch, 2006). antilock braking system The ABS control system includes ABS ECU, wheel module, brake pedal module, and hydraulic modulator (Robert Bosch, 2006). When force is applied on the brake pedal, the braking operation works through: brake booster, master cylinder, hydraulic modulator, wheel-brake cylinder and then brake. The ECU monitors the wheel speed; the master cylinder, pressure, calculates the actual and required slip and regulates the hydraulic modulator. The hydraulic modulator has solenoid valves, which adjust pressure from the master cylinder to the wheel-brake cylinder. The ABS, preventing the wheels from locking up, works when the wheel brake pressures are lowered. Figure 6. Block diagram of antilock braking system Figure 6 shows the simplified block diagram of an antilock braking system. Traction control When the driver presses the accelerator, the engine torque is converted to drive axle torque, and then is distributed to driven wheels via the transversal differential. If the increased torque can be transferred completely to the road surface, the vehicle can stably accelerate. In some low friction coefficient road surface situation, when the drive torque can not be transferred completely, the wheel will spin and the vehicle becomes unstable due to the loss of lateral stability. The traction control system (TCS) ECU and transversal differential lock controller regulates the slip of the driven wheel by determining a reference slip value, sensing the differential wheel speed, calculating distribution ratio of the transversal differential, and correcting the drive torque and braking torque (Robert Bosch, 2006). The drive torque can be adjusted by engine intervention, while the braking torque regulated for each wheel is via the expanded ABS hydraulic system. Figure 7 depicts the block diagram of a traction control system. adaptive cruise control The ACC controller includes cruise control, tracking control, and acceleration control (Robert Bosch, 2006). The cruise control adjusts the vehicle speed until it is equal to the set speed. The tracking control selects one of the vehicles ahead as the following target, and calculates the acceleration based on the distance and relative speed. The acceleration control sends the acceleration commands to the powertrain or braking system. Figure 8 demonstrates the functional block diagram of an adaptive cruise control system. 65 Vehicular Embedded System Architecture Figure 7. Block diagram of traction control system Figure 8. Block diagram of adaptive cruise control system electronic stability The electronic stability program (ESP) improves vehicle handling and braking response through programmed intervention in the braking system and/or powertrain to prevent the linear, lateral, or yaw velocity from exceeding control limits (Robert Bosch, 2006). The ESP integrates the capabilities of both ABS and TCS into a closed-loop control, and provides each wheel highly precise performance of the dynamic braking and longitudinal and lateral forces under various circumstances. 66 Figure 9 shows the simplified block diagram of an ESP control loop. The sensors involve yaw-rate, steering wheel-angle, pressure, and wheel-speed sensors, while the actuators include hydraulic modulators for braking and engine management ECU for drive torque. processors A processor is a semiconductor chip with digital circuits, which performs computational or logical tasks. It consists of three parts: the controller, Vehicular Embedded System Architecture Figure 9. Block diagram of electronic stability program system prototyping. The semicustom application-specific IC (ASIC), like gate array or standard-cell ASIC, compromises these two IC technologies mentioned above, so part of the arrays of gates or logic-level cells have already been built and system developers just need to connect these predefined components to implement their design (Vahid et al., 2002). According to the arrangement of the controller, datapath, and memory in a processor; processor technology can be categorized into the following three types: general-purpose processor, application-specific processor, and single-purpose processors. The processor technology is relevant to the hardware and software design of an embedded system, so the application is critical and they are described below. general-purpose microcontroller datapath, and memory. The controller decodes the logic of software instructions or changes the transfer among the hardware state machines. Datapath is the component executing computation or logic functions for the specified data. Memory is the storage space for data and program. Different processor technologies emphasize differently on the structure of these three processor components, so they are designed for different system architecture development. The processor is manufactured as an integrated circuit (IC) component. It is implemented in different IC technologies, which determine the level that the chip is customized. A full-custom/verylarge-scale integration (VLSI) technology optimizes performance and functions of a processor from all aspects of digital circuit design. The manufacturing of a full-custom process has high nonrecurring engineering (NRE) costs and long turnaround times. On the other hand, programmable logic device (PLD) is off-the-shelf and is ready to be used through hardware description language (HDL) programming and synthesis. It has a high unit cost, consumes more power, but it still provides reasonable performance for fast General-purpose processor or microprocessors are designed for almost all of the common applications. The logic is implemented by software programming, which has the maximum flexibility to alter its implementation, but has the lowest cost for modification. A microcontroller is a microprocessor along with some peripheral modules, like memory and I/O, integrated in a single chip without external circuits, so it provides extensive functions and is widely used in many embedded control systems. A more complicated system-onchip (SoC) technology integrates more peripheral modules in one chip, almost a complete system even including analog circuits. So the SoC microcontroller is able to implement more sophisticated functions to act as a full system. Among the processor technologies, this is the most common approach. The system implemented on a general-purpose microprocessor is referred to as a software approach. According to different applications, many manufacturers or product families, from low-end 8-bit to high-end 32-bit, can be chosen. Another issue to be considered is whether the application program is executed in an operating system (OS) or non-OS environment. 67 Vehicular Embedded System Architecture For a low-end microcontroller, the program is usually implemented in a non-OS foreground/ background mode, because the on-chip memory is small and the system clock is slow, so usually the control functions are not complicated. On the other hand, adopting a high-end 32-bit microcontroller implies the requirement of faster computations and complicated functions. Because of larger memory and a faster clock, the OS environment, which takes care of system resource management, is much better and more convenient for application development. Renesas has several microcontroller families, for instance: SuperH, H8, and M16/R32 to provide a variety of automotive functions, like powertrain, chassis, body, active/passive safety, audio and navigation. Since the microcontroller is the core of an electronic control unit and connected on the in-vehicle network, CAN or FlexRay, the network physical controller is mandatory in the microcontroller family for automotive applications. programmable logic device The system implemented on a programmable logic device is referred to as a hardware approach, because the logic is realized on synthesized logic gates. This approach is usually implemented for prototype designs and verification, and later the logic circuits will be manufactured as a full-custom IC. The throughput, power consumption, and cost of PLD are usually inferior or more expensive than that of a full-custom IC, but it is convenient for prototype development. For example, Xilinx provides XA products with Spartan-3 FPGA and IPs to implement image processing, video, or invehicle network solutions. digital signal processor The digital signal processor (DSP) is an application specific processor, which optimizes the datapath design, so execution of signal processing instructions has a much higher throughput. Moreover, 68 some dedicated hardware is added for specific applications, and the performance is improved. DSP is also a software approach, because the logic is implemented by software. Developments on DSP have advantages and drawbacks lying between microprocessor and PLD. Texas Instruments’ TMS320 DSP series with a fixed-point or floating-point provides solutions to automotive infotainment, vision control, and digital radio. sensors Sensors are applied as the input devices in an ECU for obtaining environmental and vehicular parameters. The advanced control of a modern automobile largely depends on the data acquired from sensors. Sensors convert a physical parameter to an electrical signal, and a signal conditioning circuit adjusts the electrical signal to the voltage range specified by the analog-to-digital converter (ADC). After the signal is converted to digital data, the processor can process it by applying a simple value comparison or a complicated digital signal processing algorithm. The advancement of electronic technology and material science pushes the innovation of sensor development. It improves the functions and decreases the cost of an automotive system, so vehicle features are greatly enhanced. Sensors play a more and more important role in the implementation of ECUs in vehicles. The average number of sensors per vehicle was 40 in 2007, and will be up to 70 in 2013, while the total sensor production will jump from 640 M to 1100 M in North America. Modern luxury cars have more than 100 sensors per vehicle, and total of 167 automotive applications were described. The applications of sensors in a vehicle can be categorized into three areas: powertrain, safety, and comfort (Fleming, 2008). Some sensors used in automobiles are described more in the following subsections. Vehicular Embedded System Architecture accelerometer The accelerometer, or g-cell, is applied for crash-detection. A microelectrical mechanical system (MEMS) accelerometer, which employs single-point sensing, reduces the dimension and cost by removing the wiring and harness (Monk et al., 2003). Airbag systems, including front airbags, side airbags, seat belt pretensioners and occupant sensing systems, become more complicated and need more sensing requirements. Airbag systems are required by government safety regulations, and the accelerometer plays an important role in the operation. are coded and transmitted to central console by radio frequency (RF) communications (Normann et al., 2003). Temperature Several temperature sensors are applied in monitoring engine, coolant, gear-box, fuel, intake-air, outside-air, passenger compartment, climate control unit parameter, and exhaust-gas in an automotive system. There are two temperature ranges to be measured. The first one covers -40 to 170°C, and the second range for exhaust temperature measurement covers -40 to 760°C (Bojarski, 2003). yaw-rate radar Yaw-rate sensor, angular rate sensor, or gyroscope, measure the rotation of a body in angle per unit of time along a specified axis. Typical applications include electronic stability program, rollover protection, and navigation. When concerning, the turning movement is monitored and is compared with the angle of the steering wheel and vehicle speed. Rollover accidents usually happen during high-speed driving, and causes serious injury or death to the driver and passengers, so the yaw-rate sensor must recognize the rollover of a vehicle and thus activating the safety systems appropriate protection (Schatz et al., 2003). Radar is employed for collision avoidance by finding the distance to the preceding vehicle, and ACC can be implemented by utilizing the information of distance and relative speed. Since the vehicle speed is high and response time is short on highway, a long-range radar (LRR), up to 120 m, which takes 3.6 s to collide with the stationary object at the speed 120 km/hr, is required for ACC. Short-range radars (SRR), up to 20 m, are applied for both sides of vehicle to create a safety shield around the car (Knoll, 2003). Figure 10 demonstrates surround sensing using radar and video. pressure Video A variety of pressure sensors are employed for automotive applications. Vaporized gasoline leak detection is in the low-pressure field (Yokomori et al., 2003). Suspension pressure detection and air-conditioner refrigerant pressure detection are in the high-pressure field. While gasoline fuel injection is in the very high-pressure field (Gerbers, 2003). Tire-pressure sensors measure the air pressure and temperature inside the tire. The data along with tire ID and battery lifetime Though radar can easily find the distance of an object in front, it has a narrow beam and cannot recognize objects or detect borders. A camera system is used for several applications in automotive system, like lane departure warning, night-vision improvement, object detection, blind spot warning, and distance warning. The video solution seems more feasible, but some issues need to be addressed. The captured images are essentially affected by light intensity, 69 Vehicular Embedded System Architecture Figure 10. Surround sensing using radar and video so an intelligent high-quality camera is required for automatic adjustment. A far infrared (FIR) technology is used for night-vision improvement. Real-time image processing needs intensive computing power, so a high-speed industrial computer, which may be expensive and occupies large space, is required (Knoll, 2003). wheel-speed The wheel-speed sensor measures the movement and circumstance of the tire. The ABS and TCS, prevents wheel locking and spinning, needs to know the wheel speed when applying the brake or accelerator, respectively. ESP for vehicle stability also needs wheel speed information. Other applications include transmission control, odometer, navigation system, stop and go, and roll over protection (Morbe, 2003). steering-angle Steering angle sensor is applied for ESP, which requires steering angle along with yaw rate, wheel speed, and lateral acceleration to determine if the vehicle oversteers or understeers. The sensor is also used in electric power steering, adaptive cruise control, forward intelligent lighting, navi- 70 gation system, active front and rear steering, and steer-by-wire (Kofink, 2003). Torque The torque sensors measure the powertrain torque signals on engine and transmission control units, which are part of the applications of TCS and ESP. Electric power steering is another application of torque sensors, which measure angular displacement, which is proportional to torque (Morbe, 2003). chemical There are several types of chemical sensors, including oxygen detection for air/fuel ratio control (Riegel et al., 2003), NOx sensors for emission control (Schmitt, 2003) and liquid media sensor (Jakoby et al., 2003). The exhaust gas of a vehicle has to meet emission limits, so a catalyst system is installed to do the conversion. While oxygen and NOx sensors are used for after-treatment sensing, liquid sensors are used to detect the quality of liquids, such as engine oil, gear-box oil, fuel, and battery liquid to check if they should be changed or are in abnormal states. Vehicular Embedded System Architecture acTuaTors ignition Actuators are output devices to drive electromechanical components. They are either implemented by a solenoid, which is controlled by duty cycle or PWM approach, or an electric motor, like the stepper motor or dc motor (Bonnick, 2001). Some example applications of actuators are fuel injectors, exhaust gas recirculation, ignition, ABS modulators, variable valve timing (VVT), and electric motors for hybrid/electric vehicles. The actuators of the ignition system are the combination of the spark plug, the ignition coil, and driver electronic circuits. When the controller sends a signal turning on the driver, i.e. a power transistor is conductive, current flows through the primary coil, creating a relatively large magnetic field linked to the secondary coil. The controller instantly switches off the signal, causing the transistor to be nonconducting. The sudden rapid drop in the magnetic field of the secondary coil generates a very high voltage creating the spark across the spark plug electrodes, igniting the mixture and, finally, initiating the power stroke for the engine (Bonnick, 2001). Fuel injectors Fuel injectors are electrically driven actuators that regulate the flow of fuel into an engine for engine control applications. A fuel injector is a solenoidoperated valve, which opens or closes to permit or block fuel flow to the engine. The valve is attached to the movable element of the solenoid and is switched by the solenoid activation. The quantity of fuel injected is proportional to the duration when the valve is opened, so the valve can be controlled by applying a pulse train with a specified duty cycle to the solenoid (Bonnick, 2001). exhaust gas recirculation Exhaust gas recirculation (EGR) is utilized to reduce NOx emissions. The engine controller determines the amount of EGR and sends a signal to EGR actuator. Typically, this actuator is a variableposition solenoid controlled valve that regulates the EGR as a function of intake manifold pressure and exhaust gas pressure. When the EGR valve is open, exhaust gas flows into the intake manifold. Like the fuel injector, by sending a pulse train with specified duty cycle to the solenoid, EGR actuator regulates the pressure to open the valve and then the amount of EGR (Bonnick, 2001). abs modulators Anti-lock braking system modulator contains a pump driven by an electric motor and various solenoid-operated valves. Solenoid valves with two hydraulic connections and two valve positions are used. At different degrees of brake slip, the solenoid valves switch to different pressure settings to change the pressure in the brake (Bonnick, 2001). sTorage deVices Storage devices were not commonly used in conventional vehicle systems, because there were few applications involving the storage of large amounts of data. One application for drivers or passengers to have some entertaining activities during their drive is playing audio or video from CDs or DVDs. Today, the navigation systems are more affordable and popular for eliminating the map-reading, so a mass storage device for saving the data of maps and landmark pictures is necessary. More applications for an infotainment system are developed, while the mass storage devices are mandatory for storing the multimedia contents. Hard disks are the most common storage devices in personal computers, but they don’t work 71 Vehicular Embedded System Architecture in an automotive system, which is operated in an environment with severe vibration. After the solid-state storage technology is more established and the cost is lower, the automotive applications with mass storage devices are possible. Flash memory is based on the solid-state technology and two types of flash memory, NAND flash and NOR flash, have emerged as the dominant varieties of non-volatile semiconductor memories utilized in embedded systems. The characteristics of NAND flash are high density, medium read speed, high write speed, high erase speed, and an indirect or I/O like access, so it is low cost and has been used primarily as a removable high-density data storage medium, which is appropriate for mass storage applications. The characteristics of NOR flash are lower density, high read speed, slow write speed, slow erase speed, and a random access interface, so it has typically been used for code storage and direct execution in portable electronics devices. Flash memory will increase in demand, since it can store not only multimedia data and navigation information, but also the run-time vehicle data and image as an event data recorder. human machine inTerFaces Most ECUs don’t interact with the users directly, so driver and passengers need to control the vehicle through either dashboard or other mechanical devices, like the steering wheel, brake pedal, or button/knob. Though the electronic components being used are increasing, the conventional powertrain and brake operation will not change in the near future, as a result of driving habits and safety reasons. Steer-by-wire and brake-by-wire haven’t substituted the steering wheel and brake pedal yet. However, the user interface to enhance the convenient infotainment management and comfort adjustment will be realized and popular. The human machine interface involves input and output. The conventional needle instrument 72 and button/knob will be replaced by liquid crystal display (LCD) and touch screen in the central console. Due to the limited space on dashboards and the increased amount of information that needs to be displayed, the graphic LCD display will integrate a variety of information modes into one module and can show the information flexibly. According to the usage priority or user interaction, the LCD display can show driver functions and vehicle status, navigation information, and driving assisted images. To simplify the central console appearance, the LCD display and touch screen panel are integrated into one module. The touch screen panel is mandatory in mobile devices and eliminates the usage of keyboard and mouse in a conventional PC environment. This intuitively enhances the human machine interaction on a more user-friendly graphical presentation. There are two common types of touch screen: capacitive and resistive. For security applications, biometric system and radio-frequency ID (RFID) devices are input modules for identity confirmation and personalized operation, so they enhance the convenient features of automobiles. The biometric mechanisms, including fingerprint identification, face recognition, iris recognition and voice recognition, verify if the person’s unique biometric characteristics is matching with the enrollment setting (Robert Bosch, 2006). These approaches have been researched for a variety of security applications, but most of them still do not reach a perfect successful rate in various environments or don’t have a straightforward and friendly operation process. Fingerprint identification is a more reliable approach and is employed for personalized adjustment functions and configurable favorite settings. RFID devices, using radio frequency to read and verify the identification and information, have the same application as the fingerprint approach. This technology or similar ones will be the innovative alternative to the car key and be a personal information storage device in the future. Vehicular Embedded System Architecture reFerences Bojarski, A., & Fichte, W. (2003). Temperature sensors. In J. Marek, H.P. Trah, Y. Suzuki,& I. Yokomori, (Ed.), Sensors Application Vol. 4 - Sensors for Automotive Technology (pp. 343-359). Berlin: Wiley-VCH. Morbe, M., & Zwiener, G. (2003). Whell-speed sensors. In J. Marek, H.P. Trah, Y. Suzuki,& I. Yokomori, (Ed.), Sensors application Vol. 4 Sensors for automotive technology (pp. 403-415). Berlin: Wiley-VCH. Bonnick, A. (2001). Automotive computer controlled systems. London: Butterworth Heinemann. Normann, N., Schulze, G., & Keller, W. (2003). Tire-pressure sensors. In J. Marek, H.P. Trah, Y. Suzuki,& I. Yokomori, (Ed.), Sensors application Vol. 4 - Sensors for automotive technology (pp. 536-544). Berlin: Wiley-VCH. Fleming, W. J. (2008). New automotive sensors−A review. IEEE Sensors Journal, 8(11), 1900–1921. doi:10.1109/JSEN.2008.2006452 Paret, D. (2007). Multiplexed networks for embedded systems – CAN, LIN, FlexRay, Safe-by-Wire. Hoboken, NJ: Wiley. Gerbers, J. (2003). High-pressure sensors. In J. Marek, H.P. Trah, Y. Suzuki,& I. Yokomori, (Ed.), Sensors Application Vol. 4 - Sensors for Automotive Technology (pp. 333-342). Berlin: Wiley-VCH. Riegel, J., Wiedenmann, H., & Neumann, H. (2003). Chemical sensors for oxygen detection and air/fuel ratio contol. In J. Marek, H.P. Trah, Y. Suzuki,& I. Yokomori, (Ed.), Sensors application Vol. 4 - Sensors for automotive technology (pp. 480-499). Berlin: Wiley-VCH. Jakoby, B., & Herrmann, F. (2003). Chemical sensors for liquid media. In J. Marek, H.P. Trah, Y. Suzuki,& I. Yokomori, (Ed.), Sensors Application Vol. 4 - Sensors for Automotive Technology (pp. 516-526). Berlin: Wiley-VCH. Knoll, P. (2003). Radar sensors. In J. Marek, H.P. Trah, Y. Suzuki,& I. Yokomori, (Ed.), Sensors application Vol. 4 - Sensors for automotive technology (pp. 372-385). Berlin: Wiley-VCH. Knoll, P. (2003). Video sensors. In J. Marek, H.P. Trah, Y. Suzuki,& I. Yokomori, (Ed.), Sensors application Vol. 4 - Sensors for automotive technology (pp. 386-402). Berlin: Wiley-VCH. Kofink, P. (2003). Steering-angle sensors. In J. Marek, H.P. Trah, Y. Suzuki,& I. Yokomori, (Ed.), Sensors application Vol. 4 - Sensors for automotive technology (pp. 428-449). Berlin: Wiley-VCH. Monk, D., Mladenovic, D., & Skaw, M. (2003). Accelerometers for automotive applications. In J. Marek, H.P. Trah, Y. Suzuki,& I. Yokomori, (Ed.), Sensors application Vol. 4 - Sensors for automotive technology (pp. 269-296). Berlin: Wiley-VCH. Robert Bosch Gmb, H. (2006). Safety, comfort and convenience systems. Hoboken, NJ: Wiley. Schatz, O. (2003). Yaw-rate sensors. In J. Marek, H.P. Trah, Y. Suzuki,& I. Yokomori, (Ed.), Sensors application Vol. 4 - Sensors for automotive technology (pp. 297-313). Berlin: Wiley-VCH. Schmitt, D. (2003). Chemical sensors for emission control. In J. Marek, H.P. Trah, Y. Suzuki,& I. Yokomori, (Ed.), Sensors application Vol. 4 Sensors for automotive technology (pp. 500-508). Berlin: Wiley-VCH. Vahid, F., & Givargis, T. (2002). Embedded system design: A unified hardware/software introduction. Hoboken, NJ: Wiley. Yokomori, I., & Suzuki, Y. (2003). Pressure sensors. In J. Marek, H.P. Trah, Y. Suzuki,& I. Yokomori, (Ed.), Sensors application Vol. 4 - Sensors for automotive technology (pp. 314-332). Berlin: Wiley-VCH. 73 74 Chapter 5 Data Communications Inside Vehicular Environments Cheng-Min Lin Nan Kai University of Technology, Taiwan, R.O.C. Tzong-Jye Liu Feng Chia University, Taiwan, R.O.C. absTracT ZigBee is based on IEEE 802.15.4 which specifies the physical layer and medium access control (MAC) for low-cost and low-power LR-WPAN. The technology can be applied in intelligent key, A/C operation and steering wheel inside vehicles. There are two types of devices in ZigBee, FFD and RFD. A FFD can communicate with RFDs and other FFDs, while a RFD can only communicate with a FFD. In ZigBee physical layer, it follows IEEE 802.15.4 standard and operates in unlicensed RF worldwide (2.4GHz global, 915MHz Americas or 868 MHz Europe). A superframe contained an active portion and an inactive portion is used in the MAC layer of ZigBee. The active portion includes CAP and CFP. In the inactive partition, the coordinator can enter sleep mode to save its power. Three main topologies of ZigBee are star, mesh, and tree. However, ZigBee is successfully produced into a low-cost controller applied for automotive applications, including vehicle control and status monitoring. According to the forecast of ON World in 2005 (ON WORLD, 2009), the deployed wireless sensing network nodes will increase to 127 million in 2010 from 1.2 million in 2005. It can be applied in home automation, battlefield surveillance, health care applications and vehicular environments. A wireless sensor network (WSN) constitutes a lot of wireless sensing nodes. In addition, a node in WSN consists of one or more sensors, a radio transceiver, and a microcontroller. The sensor can be used for sensing temperature, pressure, sound, vibration, motion or position, etc. to collect status from devices or environments. The transceiver is used to relay the information of the collected status computed by the microcontroller to a center node, called a gateway or sink. Therefore, a WSN belongs to one type of wireless ad-hoc networks. However, the nodes in a WSN are usually smaller than that in traditional wireless ad-hoc networks regarding node size, computing power, memory size, and transmission rage. In other words, the transmission ability, computing power, and memory size of WSN nodes are limited. DOI: 10.4018/978-1-60566-840-6.ch005 Copyright © 2010, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global is prohibited. Data Communications Inside Vehicular Environments inTroducTion In vehicular environments, a wireless personal area network (WPAN) is required because a driver is a master for the vehicle. The driver interacts with his/her vehicle according to the collected information from WPAN and a controller area network (CAN). The standard of WPANs is defined in the 15th working group of the IEEE 802.15. There are six task groups in the working group as shown in the following. • • • • • • IEEE 802.15.1: Bluetooth IEEE 802.15.2: Coexistence IEEE 802.15.3: High rate WPANs (HRWPAN), UWB IEEE 802.15.4: Low rate WPANs (LRWPAN), ZigBee IEEE 802.15.5: Mesh network IEEE 802.15.6: Body area network technologies In this chapter, we will focus on the introduction of ZigBee based on IEEE 802.15.4, a standard completed in May 2003 which specifies the physical layer and medium access control (MAC) for low-cost and low-power LR-WPAN. Although ZigBee-style networks created by the Firefly Working Group in 1999 become ZigBee later, the group does not exist now. Today’s ZigBee was adopted in 2003 and built on the IEEE 802.15.4 LR-WPAN standard and the ZigBee Alliance ratified the first ZigBee standard in December 2004 (Geer, 2005). oVerView oF ieee 802.15.4 The IEEE802.15.4 standard is the basis for the ZigBee specification. It specified the physical layer and MAC layer for LR-WPAN, providing the fundamental lower layers of WPAN. Upper layers of the protocol stack include application profiles defined by ZigBee Alliance. The architecture ZigBee/802.15.4 is shown in Figure 1. The architecture focuses on low-cost, lowspeed ubiquitous communication between devices. Furthermore, the features of IEEE 802.15.4 are illustrated as follows according to the mention of LR-WPAN Task Group (The IEEE 802.15.4 WPAN Task Group, 2009). Its applications are shown in Figure 2. • • • • • Data rates of 250 kbps, 40 kbps, and 20 kbps Star or peer-to-peer operation Dynamic device addressing Two addressing modes are implemented, including 16-bit short and 64-bit IEEE addressing Support for critical latency devices, such as joysticks Figure 1. ZigBee/802.15.4 architecture 75 Data Communications Inside Vehicular Environments Figure 2. ZigBee applications • • • • • Carrier Sense Multiple Access/Collision Avoidance (CSMA/CA) channel access is provided Automatic network is established by the coordinator Fully handshaked protocol for reliability of transmission Power management is implemented to ensure low power consumption 27 different channels, including 16 channels in the 2.4GHz ISM band, 10 channels in the 915MHz band and one channel in the 868MHz band Figure 3. FFD vs. RFD 76 Two different types of devices are defined in an LR-WPAN, a full function device (FFD) and a reduced function device (RFD). A FFD can communicate with RFDs and other FFDs, while a RFD can only communicate with a FFD. FFD can operate in the device, coordinator, and PAN coordinator modes, while RFD can only operate in the device mode. The comparison between FFD and RFD is shown in Figure 3. The ZigBee Alliance focuses on the network layer to the application layer. As shown in Figure 4, a ZigBee Device Object (ZDO) is a special device in a ZigBee network. It is responsible for a number of tasks including keeping of device roles, management of requests to join a network, Data Communications Inside Vehicular Environments Figure 4. Outline of the ZigBee stack architecture device discovery, and security. A ZDO contains ZDO Management Plane, and also defines the role of the device in the network such as ZigBee coordinator or end device. A ZDO can discover devices on the network and determine which application services they provide. ZDOs also can initialize or respond to binding requests. The last function of ZDOs is to establish secure relationship, among network devices. The functionality of Application Support Sublayer (APS) is to provide interfaces between application layer and network layer. The services of the interface are offered by two entities: data service and management. The APS data entity (APSDE) provides the data transmission service, while the APS management entity (APSME) provides the management service. The application frame is the environment whose application objects are hosted in the ZigBee device. The application objects are responsible for sending and receiving the data and provide several functions. The first function is to control and manage the protocol layers. The second function is the initialization of standard network functions. The Security Service Provider is responsible for key establishment, key transport, frame protection, and device management. Zigbee physical layer ZigBee is the technology of a novel WSN that is designed for the automation and controlling buildings, consumer electronics, and intra-vehicle wireless communications. The PHY functionalities of ZigBee are listed below. • • • • • • Activation and deactivation of radio transceiver Energy detection within current channel Link quality indication for received packets Clear channel assessment for CSMA/CA Channel frequency selection Data transmission and reception ZigBee follows IEEE 802.15.4 standard and operates in unlicensed RF worldwide (2.4GHz global, 915MHz Americas or 868 MHz Europe). There are 27 channels allocated in ZigBee standards as shown in Figure 5. Channel 0 uses the frequency at 868.0 ~ 868.6 MHz, while the data rate is 20 kbps. Channels 1 ~ 10 use the frequency at 902.0 ~ 928 MHz, where each channel can provide 40 kbps data rate. Channels 11 ~ 26 use the frequency at 2.4 ~ 2.4835 GHz, where each channel is 250 kbps. The physical layer of 868/915 MHz and 2.4 GHz uses Direct Sequence Spread Spectrum 77 Data Communications Inside Vehicular Environments Figure 5. ZigBee channel overview Figure 6. PHY frame structure (DSSS) to carry data to the channels. In 868/915 physical layer, the packet error rate must be equal to or smaller than 1% so that the hardware receiver sensitivity is required to be equal to or higher than -92 dBm. In 2.4 GHz physical layer, hardware receiver sensitivity is required to be equal to or higher than -85 dBm. The transmission power of ZigBee must be higher than -3 dBm (0.5 mW). During normal operations, the transmission power is 0 dBm (1 mW). When the transmission power is 0 dBm, the transmission range is about 10 ~ 20 meters. For variant applications, the transmission power can be adjusted so that the transmission range can be increased to 30 ~ 75 meters. The PHY frame structure of ZigBee is shown in Figure 6. There are three fields in the packet. The preamble (32 bits) is used for synchronization and the start of packet delimiter (8 bits) shall be formatted as “11100101”. The PHY header (8 bits) contains the length (0 to 127 bytes) of PSDU data field. 78 Zigbee mac layer In the ZigBee specification, the superframe format is shown in Figure 7. It is defined by the network coordinator. The superframe duration is the beacon interval broadcasted by the network coordinator. A superframe contains an active portion and an inactive portion. The active portion can be divided into 16 time slots, which can be cataloged as contention access period (CAP) and contention free period (CFP). The coordinator only receives and sends data with other devices in WPAN at active portion. In the inactive partition, the coordinator can enter sleep mode to save its power. The beacon message is broadcasted at time slot 0. The purposes of the beacon message are listed as follows: • • • • Starting superframes Synchronizing with associated devices Announcing the existence of a PAN Informing pending data in coordinators Data Communications Inside Vehicular Environments Figure 7. The superframe structure Any device wants to communicate with the coordinator in the contention access period must use a slotted CSMA/CA mechanism to access the time slots. For some applications that devices require low delay time or fixed transmission rate, the coordinator can assign some Guarantee Time Slots (GTS) for them to use. The contention free period consists of several GTSs. The maximum number of GTS is limited to seven according to the standard. This means that only the limited number of nodes can use GTS. In ZigBee, there are three data transfer models: “device to coordinator,” “coordinator to device,” and “device to device.” The data transfer models are discussed below. Device to coordinator. In a beacon-enable network, devices would find the beacon to synchronize to the superframe structure, and then they would use slotted CSMA/CA to transmit data. In a non beacon-enabled network, devices simply transmit their data using unslotted CSMA/CA. The procedure is shown in Figure 8. Coordinator to device. In a beacon-enabled network, the coordinator indicates in the beacon that the data is pending. Device periodically listens to the beacon and transmits a MAC command request using slotted CSMA/CA if necessary. In a non-beacon-enabled network, a device transmits a MAC command request using unslotted CSMA/ CA. If the coordinator has its pending data, the coordinator transmits data frame using unslotted CSMA/CA. Otherwise, the coordinator would transmit a data frame with zero length payload (Figure 9). Device to device. In a peer-to-peer topology, devices may directly communicate with other devices in the transmission range. In order to transmission data efficiently, the device transmit- Figure 8. Data transfer model (Device to coordinator) 79 Data Communications Inside Vehicular Environments Figure 9. Data transfer model (Coordinator to device) ting data can not enter the sleep mode. It would use unslotted CSMA/CA to transmit data. • Zigbee network layer The specification of ZigBee provides three types of topologies: star topology, mesh topology, and cluster tree topology as shown in Figure 10. According to the ability of ZigBee devices, we can divide them into FFDs and RFDs. FFDs have a lot of resources that include the computing capability, memory, and the power than RFDs. In addition, these topologies are made up by three types of devices. First of all, the most important type of devices is the ZigBee coordinator (ZC). Any topology can have only one ZC. Furthermore, the ZC in a ZigBee-based network is also a FFD. The ZC is responsible for network formation and maintenance. The second type of devices is the ZigBee router (ZR). The ZR is a FFD or a RFD. However, the resource in a RFD has less than that in a FFD. Moreover, the router is responsible for forwarding packets in the network. The last type of devices is the ZigBee end device (ZED). The ZED is the RFD and it cannot forward packets. In other words, the ZED cannot relay data from other devices and only talk to their parent devices. 80 • To support star topology and cluster-tree topology, the routing protocol of ZigBee uses the concept of tree routing. When a device receives a packet, it first checks if the device itself or one of its child end devices is the destination. If so, this device will accept the packet or forward this packet to the designated child. Otherwise, it would relay the packet along the tree. To support mesh topology, the ZigBee routing protocol uses the concept of Adhoc On-demand Distance Vector (AODV) (Perkins et al., 1999), an on-demand approach for finding routes. Links with lower cost will be chosen into the routing path. The cost of a link is defined based on the packet delivery probability on that link. Route discovery procedure was discussed as follows. The source broadcasts a route request packet. Intermediate nodes will rebroadcast route request if they have routing discovery table capacities and the cost is lower. Otherwise, nodes will relay the request along the tree. The destination will choose the routing path with the lowest cost and then send a route reply. The advantages of AODV are no extra traffic for Data Communications Inside Vehicular Environments Figure 10. Three topologies of ZigBee communication along existing links, and supporting unicast/ multicast. When a new ZigBee device joins the network, it has to choose a new suitable parent node. In ZigBee specification, the new parent must satisfy three conditions. First, its information is in the new ZigBee device’s neighbor table. Second, the link cost between the new ZigBee device and the parent node must be more than 3. Finally, the new parent node must have the minimum depth from coordinator in the neighbor table. In ZigBee, network addresses are assigned to devices via a distributed address assignment scheme. When the network was formed, the ZigBee coordinator would determine three network parameters. • • • Cm: Maximum number of children of a ZigBee router Rm: Maximum number of child routers of a parent node Lm: Depth of the network ZigBee defines Cm“≥ Rm, so that at least Cm“-Rm ZigBee devices can connect to a ZigBee router. A parent device utilizes Cm, Rm, and Lm to compute a parameter called Cskip, which is used to compute the size of its children’s address pools. The Cskip at depth d is computed as follows. ì ï 1 + C m × (Lm - d - 1), ï ï L -d -1 C d) = ï í ( 1 + C m - Rm - C m × Rm m skip ï , ï ï 1 - Rm ï î if Rm = 1 (a ) otherwise(b ) Address assignment starts from the ZigBee coordinator. It would first see its address and depth to 0. If a parent node at depth d has an address Aparent the nth child router is assigned t o a d d r e s s Aparent + (n - 1) * C (d ) + 1 skip th and n child end device is assigned to address Aparent + Rm * C (d ) + n . An example of skip ZigBee network address assignment was shown in Figure 11. In this example, the ZigBee coordinator assigns Cm = 6, Rm = 4, and Lm = 3. The Cskip value can be calculated by above equation and the result is 31. The addresses of the first to third child router of the coordinator were 0 + (11)31+1=1, 0 + (2-1)31 + 1 = 32 and 0 + (3-1)31 + 1 = 63. The addresses of the two devices of the coordinator were 0 + 431 + 1 = 125 and 0 + 431 + 2 = 126. The ZigBee Cluster Label K.K. Lee et al. used ZiCL (Lee et al., 2006) to improve the shortcoming of the AODV (Perkins et al., 1999) that has higher routing overhead produced in route discovery phase. ZiCL divides the ZigBee topology into several logical clusters. A cluster consists of a clusterhead and some cluster members. A clusterhead represents a cluster and 81 Data Communications Inside Vehicular Environments Figure 11. An example for address assignment cluster members are grouped into a cluster with a clusterhead. Each cluster has a unique cluster label that is assigned to each clusterhead. Some rules must be confirmed for forming a cluster: • • • • • Only the coordinator or routers can generate a logic cluster. The coordinator must be a clusterhead. Routers with even depth value can be clusterheads. Routers with odd depth value would join their parent’s cluster. End devices would join their parent’s cluster. To follow the rules, the network will form several clusters. Figure 12 shows a Cluster Labeled network. Each router node has a routing table (RT) and a route discovery table (RDT). The address information of Cluster Label is stored in RT and route discovery entries are stored in RDT for the original destination node address. A 16-bits address in ZigBee network is assigned by the ZigBee specification (Zigbee Alliance, 2006). In Figure 82 12, node Q will send data to node J that node Q is the source node and node J is the destination node. First, node Q sends a RREQ packet with the destination address of node J. The neighbor nodes of node Q receive an RREQ packet and rebroadcast it to the network until the RREQ packet reaches to node J. When intermediate nodes receive the RREQ packet, they would add an entry for the Cluster Label 0x0002 in their RT and add an entry for destination address 0x002C in their RDT. After node J receives the RREQ packet, node J would forward the RREP packet along with the reverse path created by the RREQ packet. The intermediate node H does not add the Cluster Label in its RT because the address of the node H is equal to the destination node’s Cluster Label 0x0002. Node H can send data packets through the intra-routing. When node B and node A receive the RREP packet, they make their routing entries active and forward the RREP packet to node F. Node F receives the RREP packet and it has information of adjacent node Q. Node F sends the RREP to node Q. Finally, node Q establishes a route for Cluster Label 0x0002. At the same Data Communications Inside Vehicular Environments Figure 12. The ZiCL network time, node Q sends a RNOT packet to notify its clusterhead G about the Cluster Label 0x0002 information. Node G gets the RNOT packet then it uses the Cluster Label Broadcasting to broadcast the RUPT packet to share the routing information with node T and node S. Node T and node S can maintain the routing entry for the Cluster Label 0x0002. If node T or node S has child nodes that want to send data to the destination node J, the routing path will not be established. The child nodes of the node T or node S can use the cluster label information in the RT. The nodes can save power and send data packet faster. Shortcut Tree Routing in ZigBee Networks In the cluster-tree topology of the ZigBee network, the routing path used by a tree-based routing protocol may be too long. If the destination node is near the source node in the cluster-tree network, the source node must pass packets to its node parent to the destination node. Figure 13 (a) is an example of the tree routing. The source node sends the data packet to destination. The data packet from the source node goes up to root node following the parent node, and goes down to the destination node. The data packet needs 4 hops to reach the destination node. Taehong Kim et al. proposed the shortcut tree routing (Kim et al., 2007) to improve the shortcomings of the ZigBee tree routing. Figure 13 (b) shows the concept of the shortcut tree routing, where the source node can send data packets directly to the destination node. In this way, just one hop is required to reach the destination node. The shortcut tree routing can overcome the routing overhead of the tree routing algorithm. The shortcut tree routing algorithm basically follows the ZigBee tree routing algorithm with the neighbor table which is defined in ZigBee specification. The shortcut tree routing algorithm first chooses a node from the neighbor table to be the next hop node. It computes the remaining hop counts from the next node to the destination for all the neighbor nodes, including parent and child nodes. In the neighbor table, it would choose the next hop node that can reduce the routing cost. In Figure 13 (b), it shows that all of the neighbor nodes of the remaining hops to destination node were computed. In Figure 13 (b), the source node transmits data packets to the destination node di- 83 Data Communications Inside Vehicular Environments Figure 13. Routing protocols rectly if the route cost is minimized. The ZigBee tree routing always transmits data packets to the parent node. However, the shortcut tree routing can be used to find a low cost route to save power and efficiently transmit data packets. applicaTions oF Zigbee IEEE 802.15 standards have defined three protocols of low cost wireless communications, including Bluetooth (802.15.1), UWB (802.15.3), and ZigBee (802.15.4) that can be used to perform automotive functions inside a vehicle. The comparisons among them are listed in Table 1. These wireless technologies can be applied in entertainment devices, handsets, intelligent key, A/C operation and steering wheel inside vehicles. Bickel surveyed state-of-the-art technology in wireless communication technology within vehicles, as well as between vehicles (Bickel et al., 2006), and pointed out that although information, telematics, and mobile commerce are the most prominent application areas for Bluetooth, WiFi, and WiMax, the fact that there are many different technologies of wireless networks in vehicles 84 should stimulate the market’s growth when the market matures. Hence, these technologies, such as ZigBee, are offering automakers and their suppliers many possibilities to enhance the potential of their products in the future. Next, we will focus on the technology of ZigBee for in-vehicular environments. Tsai et al. (2007) reported the results of a ZigBee-based case study conducted in a vehicle. According to their results of the experiments and measurements, ZigBee was illustrated to be a viable and promising technology for implementing an intra-car wireless sensor network. According to Table 5.2, Bluetooth for voice transmission is better than ZigBee due to a lager link bandwidth provided. However, the simulation results proposed by Wang et al. demonstrate that ZigBee can support a limited range of voice services (Wang et al., 2008). From their experimental results, two directly connected ZigBee nodes can support up to three voice over IP (VoIP) and 17 half-duplex push-to-talk (PTT) conversations. Data Communications Inside Vehicular Environments Table 1. Comparison among Bluetooth, ZigBee, and UWB (Akingbehin et al, 2005) Range Bluetooth ZigBee UWB 10m 10m <10m Chip Price $5 $2 $1 Data Rate Medium Low High Throughput Medium Low High Interference Good Good Excellent Media Voice/Data Data Video/Radar SIG Consortium Alliance Forum Data Payload 2744 104 Evolving TxPower 1 mW < 1 mW 200 µW Mode FHSS DSSS DS, MBOA Frequency 2.4 GHz .8, .9, 2.4 G 3.1-10.6 G Channels 23 or 79 1, 10, or 16 Evolving Error Correct 8-bit, 16-bit 16 CRC Evolving Security Good Good Excellent Topology Star Star, Mesh, Tree Peer-to-Peer # of nodes 7, or more 65534 Evolving Link BW 1 MHz 20-250 KHz 120M-1GHz implemenTaTion oF Zigbee A transceiver is a combination of transmitter and receiver in a single package. In a radio transceiver, the receiver is silenced while transmitting. An electronic switch allows the transmitter and receiver to be connected to the same antenna, and prevents the transmitter output from damaging of the receiver. ZigBee chip vendors typically sell integrated radios and microcontrollers with 60K ~ 128K bytes flash memory, such as the Maxstream Xbee, the MeshNetics ZigBit and the CEL ZFSM-100 as shown in Figure 14. Radios were also used stand-alone with any processors or microcontrollers. Generally, the chip vendors also offer the ZigBee software stack, although independent ones are also available. The RF transceiver is just one component in a ZigBee platform solution. A processing device, such as an MCU, is required to complete the entire solution by housing the IEEE 802.15.4 MAC and ZigBee software. A typical ZigBee protocol stack implementation needs a flash memory from 64K to 96K bytes. Based on the requirement of applications, there are several special functions, such as flash memory size, ADC support, GPIO and operating voltage, need to be considered. The implemented platform is an autonomous, Figure 14. ZigBee modules 85 Data Communications Inside Vehicular Environments Figure 15. The implemented ZigBee platform block diagram lightweight wireless communication and computing platform based on MaxStream Inc. Xbee radio module and a microprocessor. The platform has no integrated sensors, since individual sensor configurations are required depending on the application. Instead, through predetermined connector, the platform can be used with various serial devices, such as digital/analog sensors and SPI compatible devices. Figure 15 shows an implemented ZigBee system block diagram, where Ateml Atmega64L is used as the system microcontroller. The Atmega64L is a Figure 16. ZigBee platform implementation 86 high performance 8 bit RISC microcontroller with 64K bytes The peripheral of Atmega64L contains two 8 bit timers, 8-channel 10-bit ADC, dual programmable serial USARTs and Master/Slave SPI serial interface. For the IEEE 802.15.4 transceiver chip selection, we used a MaxStreem incorporated product, the Xbee OEM RF module. The module of Xbee supports two operation modes:Transparent mode and Application Programming Interface (API) mode. The Xbee module operates in transparent mode defines a set of AT command for controlling Data Communications Inside Vehicular Environments Table 2. Components of implemented platform Part Type Designator Footprint Part Type Designator Footprint Part Type Designator Footprint 0.1uF C13 402 4.7uF/16V C14 0805C 6.8uH L2 C3-Y1.5R 0.1uF/16V C17 402 10uF/10V C15 0805C DB9 CON1 DB-9/M 0.1uF/16V C20 402 10uF/10V C16 0805C DCJACK J5 DCJACK 22pF/50V C23 402 104p C2 0805C 1N4148 D1 DS 400K/1% R10 402 10uF/10V C21 0805C AVR_ISP Jtag1 FKV10SN 10K R8 402 10uF/10V C22 0805C CON10 J1 IDC-10 100K/1% R9 402 104p C3 0805C CON10 J2 IDC-10 0.1uF/25V C10 0603C 0.1u/16V C4 0805C JUNPER J4 IDC-3 0.1uF/25V C11 0603C 10uF/10V C5 0805C CON6 J3 IDC-6 0.1uF/25V C18 0603C 20P C6 0805C 10uH L1 L2520 1uF/10V C19 0603C 20P C7 0805C JUMPER JP2 sip2 0.1uF/25V C8 0603C GREEN LED1 0805LED JUMPER JP3 sip2 0.1uF/25V C9 0603C GREEN LED2 0805LED MAX3232 U3 SO16 27R R11 0603R RED LED3 0805LED RT8008 U2 SOT-25 27R R12 0603R 47 R1 0805R ATMEGA64L U1 TQFP-64 0R R6 0603R 47 R2 0805R 8.00MHz XTAL1 X1 0R R7 0603R 47 R3 0805R XBEE U4 XBEE 0.1u/16V C1 0805C 10K R4 0805R 10uF/10V C12 0805C 4.7K R5 0805R the internal ZigBee protocol stack. Via the AT command, the user can quickly establish an application without the detailed knowledge of ZigBee protocol stack. In the API mode, the Xbee module acts as a pure IEEE 802.15.4 RF module. Only a serial communication protocol is defined in the API mode. Hence, the full ZigBee stack is required to be implemented in applications and the protocol is used to communicate with the Xbee module. In Figure 16, it shows a ZigBee platform implementation. The implemented system has two power sources: mains-power and battery-power. The power source can be selected from jumper 3. LED3 is a power indicator. LED2 and LED1 are connected to GPIO port C of Atmega64L. These two LEDs can be utilized for general purpose. The system reserves a 8-bit GPIO port and a 8-bit ADC connector for extensions. The SPI is also reserved for changing other IEEE 802.15.4 radio transceivers, such as AT86RF230. Table 2 lists the components of the implemented platform. Figure 17 is the platform circuit diagram which was layout by Protel 99SE. The connection between MCU and XBee is via the UART1 of MCU. Regarding the power source, there are two sources (main and battery power) that can be switched via jumper 4. summary The chapter introduced the standard of low rate WPANs to be called ZigBee. Two types of devices, FFD and RFD, are discussed. We have known that a FFD can communicate with RFDs and other FFDs, while a RFD can only communicate with a FFD. 87 Data Communications Inside Vehicular Environments Figure 17. Circuit diagram of the implemented platform In ZigBee physical layer, ZigBee follows IEEE 802.15.4 standard and operates in unlicensed RF worldwide (2.4GHz global, 915MHz Americas or 868 MHz Europe). A superframe contained an active portion and an inactive portion is used in the MAC layer of ZigBee. The active portion includes CAP and CFP. In the inactive partition, the coordinator can enter sleep mode to save its power. Three main topologies of ZigBee are star, mesh, and tree. However, ZigBee is successfully produced into a low-cost controller applied for automotive applications, including vehicle control and status monitoring. 88 reFerences Akingbehin, K. (2005). Wireless communications for intra-vehicle use (Wireless harnesses). Institute for Advanced Vehicle Systems, University of Michigan-Dearborn, Dearborn, MI. Bickel, G. S. (2006). Inter/intra-vehicle wireless communication. Retrieved from http://userfs.cec. wustl.edu/~gsb1/index.html Geer, D. (2005). Users Make a Beeline for ZigBee Sensor Technology. Computer, 16–19. Kim, T., Kim, D., Park, N., Yoo, S., & Lopez, T. S. (2007). Shortcut tree routing in ZigBee networks. In . Proceedings of ISWPC, 07, 42–47. Data Communications Inside Vehicular Environments Lee, K. K., Kim, S. H., Choi, Y. S., & Park, H. S. (2006). A mesh routing protocol using cluster label in the ZigBee network. In Proceedings of Mobile Ad-hoc and Sensor Systems (MASS), (pp. 801-806). Perkins, C. E., & Royer, E. M. (1999). Ad-hoc On-demand distance vector routing. In . Proceedings of WMCSA, 99, 90–100. The, I. E. E. E. 802.15.4 WPAN Group. (2009). Retrieved from http://www.ieee802.org/15/pub/ TG4.html 89 90 Chapter 6 Wireless Access in Vehicular Environments Tzong-Jye Liu Feng Chia University, Taiwan, R.O.C. Ching-Wen Chen Feng Chia University, Taiwan, R.O.C. absTracT The IEEE 1609 standards define communication for wireless access in vehicular environment (WAVE) services, which enable vehicle-to-vehicle, vehicle-to-roadside, as well as vehicle-to-infrastructure communications. The standard consists of four parts, which are briefly described in this chapter. IEEE 1609.1 describes the WAVE resource manager which specifies the wireless access method in a WAVE environment and allows a remote manager application to establish connection with a resource command processor on an on-board unit. IEEE 1609.2 defines several secure message formats to process messages for WAVE system. The standard covers methods for securing WAVE management messages and application messages, which protects messages from attacks such as eavesdropping, spoofing, alteration, replay, and linkable information to unauthorized parties. IEEE 1609.3 defines network services for WAVE systems, whose network services operate at the network and transport layers of the OSI model and support both the IPv6 traffics and the WAVE short message services. IEEE 1609.4 describes WAVE multi-channel operations. It specifies the functions of MAC sublayer management entity and WAVE MAC with channel coordination. The multi-channel operation provides an efficient mechanism that controls the operation of upper layer across multiple channels. inTroducTion The IEEE 1609 standards (IEEE 1609.1TM, 2006; IEEE 1609.2TM, 2006; IEEE 1609.3TM, 2007; IEEE 1609.4TM, 2006) define communication for DOI: 10.4018/978-1-60566-840-6.ch006 wireless access in vehicular environment (WAVE) services, which enable vehicle-to-vehicle, vehicleto-roadside, as well as vehicle-to-infrastructure communications. The standard consists of four parts, which are briefly described in this chapter. IEEE 1609.1 (IEEE 1609.1TM, 2006) describes the WAVE resource manager which specifies the wire- Copyright © 2010, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global is prohibited. Wireless Access in Vehicular Environments less access method in a WAVE environment and allows a remote manager application to establish connection with a resource command processor on an on-board unit. IEEE 1609.2 (IEEE 1609.2TM, 2006) defines several secure message formats to process messages for WAVE system. The standard covers methods for securing WAVE management messages and application messages, which protects messages from attacks such as eavesdropping, spoofing, alteration, replay, and linkable information to unauthorized parties. IEEE 1609.3 (IEEE 1609.3TM, 2007) defines network services for WAVE systems, whose network services operate at the network and transport layers of the OSI model and support both the IPv6 traffics and the WAVE short message services. IEEE 1609.4 (IEEE 1609.4TM, 2006) describes WAVE multi-channel operations. It specifies the functions of MAC sublayer management entity and WAVE MAC with channel coordination. The multi-channel operation provides an efficient mechanism that controls the operation of upper layer across multiple channels. This chapter describes the standards for the wireless access in vehicular environment (WAVE). The IEEE 1609 standards define communication for WAVE services, which enable vehicle to vehicle (V2V), vehicle to roadside, and vehicle to infrastructure (V2I) communication. The communication system integrates the information of engine, gearing, brake, roadside unit, and provides safety services for drivers. In the standard, the specification of the physical layer is defined in IEEE 802.11p and the communication protocol for WAVE network service is IPv6. This chapter also describes the four parts of the IEEE 1609 standards. Section 6.2 describes IEEE 1609.4 (2006), the functions of MAC sublayer management entity (MLME) and WAVE MAC with channel coordination are introduced. Section 6.3 describes the IEEE 1609.3 (2007), the WAVE network services for WAVE systems are introduced. Section 6.4 describes the IEEE 1609.2 (2006), the secure message formats and the process of the secure messages for DSRC/WAVE system are defined. Section 6.5 describes the IEEE 1609.1 (2006), the WAVE resource management is introduced. waVe mulTi-channel operaTions IEEE 1609.4 (2006) describes WAVE multichannel operations. It specifies the functions of MAC sublayer management entity (MLME) and WAVE MAC with channel coordination. Multichannel operation provides an efficient mechanism that controls the operation of upper layer across multiple channels. The channel coordination enhances the mechanism defined in the MAC of IEEE 802.11 and interacts with IEEE 802.2 logical link control and IEEE 802.11p PHY. WAVE devices (on-board units or roadside units) provide an architecture that supports a single control channel (CCH) and multiple service channels (SCHs). The control channel is for transmitting WAVE short message and announcing WAVE services. The service channels are for interactions and transmissions between applications. WAVE standard uses the specification of the PHY in IEEE 802.11 and revises to IEEE 802.11p. The services defined in IEEE 1609.4 (2006) are for managing the channel coordination and supporting MAC service data unit (MSDU) delivery. The services include the channel routing, user priority, channel coordination and MAC service data unit transfer. In the following of this section, we will describe these four services. The channel routing WAVE supports both the WAVE short message (WSM) and IP datagram transfer. When an MSDU is passed from the LLC to the MAC, the MAC determines whether the MSDU is a WSM or an IP datagram by checking the EtherType field in the MSDU. If the value of this field is 0x86DD, an IPv6 header follows. If the value is 0x88DC, 91 Wireless Access in Vehicular Environments a WAVE short message protocol (WSMP) header follows. In the following of this subsection, the channel routing for WSMP data and IP datagram is described. Channel Routing for WSMP Data As is shown in Figure 1, WSMP header contains the channel, power level and data rate associated with the data packet. The scenario of channel routing for WSMP data is as follows: First, WSMP data is passed from the LLC to the MAC. Then, MAC routes the packet to a proper queue according to the channel number contained in the WSMP header. If the channel number does not corresponding to the control channel number or the current service channel number, it is invalid. The data packet is discarded if the channel number is invalid. Routing for IP Datagram The transmitter profile must be registered to the MAC sublayer management entity (MLME) before initializing IP datagram exchanges. The transmitter profile contains a service channel number, power level, data rate and the adaptable status of power level and data rate. At any given time, only one transmitter profile may be active. Then, IP datagram is passed from the LLC to the MAC; and the MAC routes the packet to a Figure 1. Portion of WSMP header used to control PHY (IEEE 1609.4TM, 2006) 92 data buffer that corresponds to the current service channel. If there is no transmitter profile registered or the transmitting WAVE device is not a member of any WAVE Basic Service Set (discussed in 6.3.2), the datagram is discarded. user priority IEEE 802.11e Enhanced Distributed Channel Access (EDCA) mechanism is used to contend for accessing medium. The MAC buffers the data by mapping its user priority to access category index (ACI) is defined in IEEE 802.11. The general architecture of prioritized access on one channel for the data transmission is shown in Figure 2. When MAC receives a MSDU and completes the channel routing process, MAC maps its user priority to access category index. Each access category has a unique and independent channel function. The channel function is used to pick a data packet from the access category to compete access right. Then, the package being picked Figure 2. Prioritized access for data transmission on one channel. (IEEE 1609.4TM, 2006) Wireless Access in Vehicular Environments Figure 3. Sync interval, guard interval, CCH interval, SCH interval (IEEE 1609.4TM, 2006) decides which packet gets the access right according to its back-off. The packet with the smallest back-off gets the access right. • • channel coordination Now, we would describe the channel coordination function, which is done mainly according to the MAC layer synchronizing operation. Based on the synchronization operation, packets could be sent from MAC layer into wireless channel. Shown in Figure 3 is the sync interval, it contains CCH interval and SCH interval components. A buffering time interval, called Guard Interval, is used to synchronize various devices (for example, synchronization of their time). For WAVE devices, Coordinated Universal Time (UTC) is used as the reference time. The length of the sync interval is dot4SyncInterval; and sync interval starting at the instant the time synchronization function timer modulo dot4SyncInterval is zero. All WAVE devices shall monitor the CCH during the CCH interval. mac service data unit Transfer In this subsection, we describe the MAC service data unit transfer defined in IEEE 1609.4 (2006). As it was shown in Figure 4, the data transmission processing flow is as follows: • IP or WSM requests for data transmission. LLC passes an MSDU to MAC; and the channel router in MAC distinguishes between IP and WSMP by checking EtherType field. ◦ If the value of EtherType is WSMP, the channel router assigns the MSDU to the access category (AC) based on the channel number and the user priority in the WSMP header. ◦ If the value of EtherType is IP, the channel router allocates the service channel, access category and user priority to the MSDU. Then, the channel router puts the MSDU to the channel buffer according to its user priority. ◦ When the data unit wins the contention on the current channel, ◦ If the data unit is WSM, the packet is transmitted by using the transmit power and the data rate recorded in the header. ◦ If the data unit is IP datagram, the packet is transmitted by using the transmit power and the data rate stored in the registered transmitter profile. The value of the power level and the data rate is set in TXVECTOR. TXVECTOR is a set of parameters that the MAC provides 93 Wireless Access in Vehicular Environments Figure 4. Data transmission processing flow (IEEE 1609.4TM, 2006) • • • • PHY to transmit the data unit. A clear channel is reported by sending an IDLE command from PHY to MAC. MAC sends a TXVECTOR command for setting; and a confirmation is replied after PHY completes the setting. PHY completes the setting according to the power level and the data rate. Data is exchanged between PHY and MAC through a series of actions. A confirmation is replied after PHY completes the transmission. At the end, LLC receives an indication from MAC. Now, consider the data transfer in the control channel. WSMs conforming to the WSMP can be exchanged directly; and the transmission of IP datagram is not permitted. Now, let us consider the data transfer in the service channel. In order to initiate communications on a service channel, a roadside unit or an on-board unit transmits WAVE announcement action frames on the control channel to advertise offered services available on that service channel. An on-board unit receives the announcement on the control channel and generally 94 establishes communications with the provider on the specified service channel. waVe neTwork serVices IEEE 1609.3 (2007) defines network services for WAVE systems. WAVE network services operate at the network and transport layers of the OSI model, supporting a high data rate, low latency communication between WAVE devices. The WAVE network services support both the IPv6 traffics and the WAVE short message (WSM) services. overview The WAVE network services provide addressing and routing services in a WAVE system. It operates at the network and transport layers and supports wireless connectivity among vehicle-based devices. It also supports connectivity between fixed roadside devices and vehicle-based devices. The wireless communication is based on the 5.9 GHz DSRC/WAVE mode. Wireless Access in Vehicular Environments Figure 5. WAVE protocol stack (IEEE 1609.3TM, 2007) The scope of IEEE 1609.3 (2007) standard is shown in Figure 5. The WAVE system supports both IP and non-IP applications. The communications for non-IP based applications are based on the WAVE short message protocol (WSMP) and the IP based applications are based on IPv6. In IEEE 1609.3 (2007), the standard defines the management information base (MIB) of the WAVE management entity (WME). It also specifies the functions of LLC, UDP/TCP and WAVE short message protocol. The WAVE protocol stack consists of data plane and management plane. • Data plane: The data plane consists of the communication protocol and hardwares for data transfer. The data plane of the WAVE network services must support UDP protocol; and the TCP protocol is optimal. The WAVE network services shall support the IPv6 protocol and the Logical Link Control (LLC) protocol as specified in IEEE 802.2. The data plane of the WAVE network services also supports the WAVE short message. The WAVE short message can be delivered to multiple destinations. The implementation of the WAVE • short message protocol has to support the forwarding function. Management plane: The management plane of the WAVE network services provides the following services: application registration, WAVE Basic Service Set (WBSS) management, channel usage monitoring, IPv6 configuration, Received Channel Power Indicator (RCPI) monitoring, and the Management Information Base (MIB) maintenance. About the communication protocol of the WAVE system, we can go through the channel type and two WAVE-supported protocols for explanation. The WAVE-supported protocol contains both the WAVE short message protocol and standard IPv6 protocol. • Channel types: The WAVE system supports two types of channels: the control channel (CCH) and the service channel (SCH). In a WAVE system, there is only one control channel and multiple service channels.The control channel is reserved for system control message and short, but high-priority applications. The service 95 Wireless Access in Vehicular Environments • • channels support general purpose data transfer applications. WAVE short message protocol (WSMP): The WAVE short message protocol is designed for optimized operation in the WAVE environment. The WAVE short message may be sent on both the control channel and the service channels. In a WAVE system, the WAVE short message is used for message transmission. It also allows application to directly control physical layer parameters, such as the channel number and/or the transmitter power. It is designed to minimize to channel capacity. Based on the Provider Service Identifier (PSID), the WAVE short message may be sent to the correct destination. Standard Internet Protocol (IPv6): The transmission based on the standard IPv6 can be only used on the service channels (SCH). The protocol is for generic applications and network services. nel and distributes this information by using the control channel. WBSS used role-based for controlling regardless of users’ identity, restricting users’ access rights. Applications could integrate in the priority designing, for example if two applications requesting for the accessing of the same channel, WAVE management entity would give the access according to the application’s access level. The access rights of lower layer are decided and controlled by MAC layer. Data transfer by an application can be operated both with a WBSS and without a WBSS. • wave system operations In a WAVE system, a WAVE Basic Service Set (WBSS) is established to support data traffics between applications. A set of cooperating WAVE stations consists of a single WBSS provider and none or multiple WBSS users. Applications based on the WAVE short message protocol may initiate a WBSS to allocate a service channel, but this is not required since the WAVE short messages may be exchanged on the control channel. We would move on to the communication concept of WAVE. Applications themselves can decide whether WBSS would be employed. When WBSS is not used, WSMs could use only CCH. If an application exchanges data with a WBSS, it may send the information by using WAVE short message protocol or IPv6 protocol on the service channel. A WAVE device may announce a WAVE Basic Service Set on the service chan- 96 • Operation without a WBSS: Operation without a WBSS uses WSMP to exchange data on the control channel. A source application composes the WAVE short message data and addresses it to the broadcast MAC address. Then, the application selects appropriate radio channel information to control the transmission, and requests the WSMP for delivery the message. A receiving device accepts the packet and passes it up to the protocol stack. The WSMP protocol stack delivers it to the registered application(s), based on PSID. Then, the receiving application knows the address of the provider device, and may exchange data on the CCH if desired. Operation with a WBSS: There are two types of WBSS: persistent and nonpersistent. A persistent WBSS may support Internet access; and a non-persistent WBSS may support on-demand services. A persistent WBSS is announced during CCH interval and the usage of the persistent WBSS will offer an ongoing service to any devices that come into the range of WBSS. Non-persistent WBSS is announced only on initiation. The usage of the non-persistent WBSS will support an on-demand service. Wireless Access in Vehicular Environments WAVE applications offer services to potential user applications by announcing their air interface via a WAVE service information element (WSIE). WAVE service information element is inside a WAVE announcement frame. The process of creating a WSIE and transmitting it is initiated when an application request to initiate a WBSS and offer services. On receiving of the WME application request, the WME starts the WBSS. Upon receiving of the notification of the WBSS initiation from the WME, an application is free to generate data packets (WSM or IPv6 packet) for transmission on the service channel. Any received packets destined for the application will be delivered to the application via the WSMP or IPv6 stack. The WBSS remains active at the local device until it is ended. The WME would terminate a WBSS if any of the following reason holds. • • • All applications have completed their activities on the WBSS. The WBSS is no longer being used by locally-registered applications. A higher priority application participates the WBSS and induces the conflict. If the lower layer indicates that the service channel is idle, a user device may terminate its participation in the WBSS. waVe securiTy serVices IEEE establishes IEEE 1556 to develop the standard for WAVE security service. Later, the standard was renamed as IEEE 1609.2 (2006). The main goal of this standard is to define the secure message formats to process messages for DSRC/WAVE system. The standard covers methods for securing WAVE management messages and application messages. The exception of vehicle-originating safety messages and the administrative functions for the core security functions are also described. The standard protects messages from attacks such as eavesdropping, spoofing, alteration, replay, and linkable information to unauthorized parties. The standard specifies the secure message format by using a presentation language based on TLS (IETF RFC 2246). The statement of the presentation language includes variable names, data types and functions. Variable names are all lowercase. Multiple words name is indicated by underscores. Data types begin with an uppercase letter; they may contain a mixture of upper and lowercase. secured messages Many of the applications discussed in this standard use the secured message format in Figure 6. The generic message format includes the protocol version and the type. The protocol version is the current version of the protocol. The type contains the type of the message; it tells the receiver how to interrupt the received message. The type value 0 denotes unsecured, the type value 1 denotes signed, and the type value 2 denotes encrypted. Type values from 240 to 255 are reserved for private usage. The message types are as follows (IEEE 1609.2TM, 2006): • • • • SignedMessage, ToBeSignedMessage, and MessageFlags types: These types are used when the type field is signed. S i g n e d W S M & To B e S i g n e d W S M types: The ToBeSignedWSM is the variant of the ToBeSignedMessage except that the application field is omitted. The SignedWSM types is also the variant of the SignedMessage, expect that the signed data is a ToBeSignedWSM structure. PublicKey, PKAlgorithm, & SymmAlg orithm types: The PublicKey structure is for encoding a public key and identifies the algorithm that public key is used. ECPublicKey type: An ECDSA or ECIES public key is specified. 97 Wireless Access in Vehicular Environments Figure 6. Secured message format (IEEE 1609.2TM, 2006) • • • • • • 98 CertID8 & CertID10 types: These types are used to identify a certificate. The CertIDx, x is 8 or 10, for a given certificate is calculated by the SHA-256 hash function and the low-order x bytes of the hash value is output. The ApplicationID & FullySpecifiedAp pID types: The security services use these types to determine whether a message may legitimately address a given application. Time64 & Time32 types: Time64 is a 64-bit integer. It is encoded in big-endian format and gives the number of microseconds. Time32 is a 32-bit integer. It is also encoded in big-endian format and gives the number of seconds. SignerInfo type: The receiver uses SignerInfo structure to determine which keying material is used to authenticate the message. Signature type: The actual signature is contained in the signature field. ECDSASignature type: The ECDSA signatures are performed. • • • • • EncryptedMessage, EncryptedCont entInfo, & RecipientInfo types: The EncryptedMessage type is supported to encrypt a message to one or more recipients using the recipients’ public keys. ECIESNISTp256EncryptedKey & AESCCMCiphertext types: The ECIESNISTp256EncryptedKey type is used to transfer the symmetric key. The symmetric key is encrypted using ECIES as specified in IEEE Std 1363a-2004. WAVECertificate, ToBeSignedWAVECe rtificate, CertSpecificData, SubjectType, & CRLSeries types: The standard supports many certificate types. The certificate type corresponding to different roles that signers can play in the system. All these certificates use the same structure. WAVECRL,ToBeSignedCRL, CRLType, and IDAndDate types: These types are used for certificate revocation list. WAVE Certi fi ca teReq u es t & WAV ECertificateResponse types: The WAVECertificateRequest type is used to request a WAVECertificate. The Wireless Access in Vehicular Environments • • • WAVECertificateResponse is used to return a WAVECertificate. GeographicRegion & RegionType types: These types are for defining region. The 2DLocation & 3DlocationAndConfi dence types: 2Dlocation is used to define validity regions using in the certificates. 3DlocationAndConfidence is used to identify the transmission locations and includes a confidence field. Certificate scopes: This type defines the scope of the certificate. signed messages When an entity wants to create a signed message for transmission, it has to contain the following information and services: • • • • • A key for signing. The certificate associated with the signing key. A random number generator. A cryptographic implementation. The current position and time; and the estimated error in that position and time. Then, the entity signs the message by taking the following steps: 1. 2. 3. Fill the ToBeSigned Message structure and encode it as the unsigned value octet string. Sign the unsigned message value. Create and encode the signed message value. An entity may receive a signed message. The signed message contains: • • • The recently received message. The root certificate. The revoked certificate. The steps to process the signed message are as follows: 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. Decode the received octet string. If the type of the application field in the signed message is not correct, discard the message. Check and ensure that the message is not a replay message. If the transmission location is included in the signed message, perform the geographical validity checks. Use the signer digest to retrieve the certificate from the message certificate cache. Verify the message certificate. Validate the certificate chain. If needed, verify that the transmission location. Check if it is within the GeographicRegion of the message signing certificate. Verify the application field in the message consists with that in the certificate. Verify that none of the certificates in the certificate chain has been revoked. Verify the signature on each certificate with the public key from its issuing certificate. Verify the signature on the message with the public key from the message signer’s certificate. Cache any previously unseen certificates, their associated Application IDs and associated Geographic Regions. encrypted messages The processes to encrypt messages are as follows. 1. 2. 3. 4. Retrieve the certificate Check that the recipient’s certificate has not been revoked Select a symmetric encryption algorithm in the recipient’s certificate Generate a random key for this algorithm 99 Wireless Access in Vehicular Environments 5. 6. 7. 8. Retrieve and encode the message to be encrypted Encrypt the message Create a RecipientInfo field Create and encode the EncryptedMessage. If an entity wants to send a signed and encrypted message, first it creates a signed message from the original message. Then, it creates an encrypted message from the encoded SignedMessage. When an entity receipts a signed and encrypted message, first it decrypts the message. Then, it forwards the output message to the application for processing. waVe resource manager This section describes the WAVE Resource manager, which is standardized in IEEE 1609, specifying the wireless access method in a WAVE environment. The proposed access method allows a RMA (remote manager application) to establish connection with a RCP (resource command processor) on an OBU (on-board unit) through the RM (resource manager) of a RSU (roadside unit). A RM in WAVE takes the role of a multiplexer, directing the information exchange between RMA and RCP. WAVE specification aims at creating a completely interoperable communication environment through which vehicles and RSUs could transfer data effectively. wave architecture As shown in Figure 7, a RCP resides as a component in an OBU, while the RM can be a component of either the OBU or a RSU. RM and a resource manager application communicate via the security tunnel in the between. RM, as a the multiplexer in the WAVE framework, allows each RMA to perform end-to-end communicate with RCPs on OBUs through itself, while the link between RM and RCP is established through wireless (IEEE 802.11p), which might be insecure. When a RM of a RSU sends respective message to a RCP on an OBU, the command would be executed by the RCP. Should the need arise; it could response to the requesting RMA through the RM. Therefore in Figure 7, RM provides services for RMAs to access the memory and UIs on the OBU, which are controlled by RCP. As shown in Figure 7. Components addressed in IEEE 1609.1 (IEEE 1609.1TM, 2006) 100 Wireless Access in Vehicular Environments Figure 8. Data transfer in WAVE Figure 9. Four basic forms for services Figure 7, if a RM resides in an OBU, this figure thus then depicts the situation of an inter-vehicle communication session. wave data Transfer As shown in Figure 8, the data transfer of RMA is done through command and response, which itself includes a two phase action: the encapsulation of the information from RMA to RM containing in the APDU (application protocol data unit), and the command/response as a whole from RM to RCP. The steps are detailed below: 1. 2. 3. RM received commands from a RMA; after confirmation and execution, the command is send to the RCP of the OBU via IEEE 802.11p. Message from RM to RMA is encapsulated in APDU, and those from RM to RCP in UDP packets. Command executed by RCP. RCP responds the execution result as well as other information back to RMA through RM. We will then talk about the data structure of the message. When the RMA and the RM are not in the same device, messages will be delivered in packets, thus APDU; should they be in two stack protocols of a single identical device, either RSU or OBU, we would then pass it as ASDU(application service data unit). The messaging between interfaces of different protocol layers is called SAP (service access point). As shown in Figure 9, there are four types of service indicated in this specification: Request, Indication, Response, and Conform, each service can be in one of the following modes: Confirmed Mode, Non-confirmed Mode, and Locally confirmed Mode (as indicated in Figure 10(a)~(c)). application components In application layer, there are several components: OBU resources, RM commands and response, and Services provided by the RM. • OBU resources: OBU resource covers both Memory and User Interfaces: 101 Wireless Access in Vehicular Environments Figure 10.Three modes for services ◦ 102 OBU memory: OBU memory is divided into partitions, each further divided into addressable blocks called pages, which are accessed by RMA through RM command set. Since each partition is indexed by a 16-bit integer, the largest partition is 64k bytes. In the same fashion, a largest page is also 64k bytes. There must exist partition 0 in an OBU memory, while additional partitions are optional. Page 0 of partition 0 is reserved for special purpose, and all access to this page would be deemed invalid. OBU memory pages come in three types: storage pages (type 0), memory-mapped pages (type 1), and transfer pages (type 2), which are detailed below: ▪ Storage pages (type 0): Generalpurpose pages, type 0 pages are controlled by applications, providing RMA data storage and retrieval. ▪ Memory-mapped pages (type 1): OBU integrates many UI devices, which are controlled via memory-mapped pages. These pages are treated as buffers for devices. When a RMA want to control a UI device, it simply writes the data to the page of that device. The command for Wireless Access in Vehicular Environments • such action is Set User Interface command. To get data from a UI device, RMA executes Read Memory Page command. ▪ Transfer pages (type 2): Transfer pages are the interfaces between onboard equipments and the network, and are related to RCP-controlled external interfaces. When RMA writes data to a transfer page, the data entry would be sent to the interface it was related by. ◦ User interface: User interface acts as bridges between machine and men. Related GUIs are introduced below: ▪ Visual display: Through which an OBU could display message. Acceptable colors in this specification are red, green and blue. ▪ Buzzer: Through an OBU could notify user of important events. ▪ Enunciator: Through which an OBU can read messages. ▪ Character readout: Through which an OBU can display text messages. ▪ Keypad: Through which a user can input message. ▪ Other future GUI RM commands and response: Command format for RM is shown in Figure 11. Each command is formed by 5 bytes, transferred in sequence with the MSB of each byte sent first. Fields in the command are explained below: ◦ Byte 1: Bit 7 is reserved; next 7 bits form the Command ID, which is unique for each command. For instance, the ID for Read Memory Page is 0x10. Refer to IEEE 1609.1 (2006) for more information. ◦ Byte 2: Bit 7 is No Response Indicator, which, for its namesake, indicates whether the receiver should reply, TRUE for no and FALSE for yes. Following 7 bits form the Command Transaction ID, which is set by RMA when sending command to RCP. When replying, this ID will be included without modification so that RMA could identify to which message it was responded. ◦ Byte 3~4: This field is 2-byte long, used to indicate the length of the following Command parameter field. ◦ Byte 5~: Bytes after byte 5 is for the parameters which RMA sends to RCP. ◦ Commands in resource management of WAVE are passed in the form of a command sequence. When passing a single command, it would still be transferred in a sequence with a single entry. Sequences are processed in FIFO fashion once received by the client. RM in WAVE would then process each command sequence until a invalid command is met. While executing, if a FALSE No_Response_Indicator value is met, the RCP would need to send in a replying response. Shown in Figure 12 is the response format of RM: • • • • Byte 1: bit 7 is reserved, with the following 7 bits forming the Command ID. Byte 2: bit 7 is unused, with the following 7 bits forming the Command Transaction ID. Byte 3: this is the response status of RCP, for example Command Success (0x01), Command Failed (0x02), refer to IEEE 1609.1 for more info. Byte 4~5: This field is 2-byte long, used to indicate the length of the following Response Data. Note that this field and the 103 Wireless Access in Vehicular Environments Figure 11. Command format. (IEEE 1609.1TM, 2006) • following one are used only when making reply for Read Memory Page command. Byte 6~: Response Data Following IDs are further specified in IEEE 1609.1 (2006): • • • • • • Read Memory Page command (ID: 0x10) Write Memory Page command (ID: 0x11) Insert Message command (ID: 0x12) Sleep Transaction command (ID: 0x30) Reserve Memory Page command (ID: 0x40) Release Memory Page command (ID: 0x41) Figure 12. Response format. (IEEE 1609.1TM, 2006) 104 • • Reserve Partition command (ID: 0x43) Release Partition command (ID: 0x44) • Services provided by the RM: RM takes the role of the application layer in WAVE by offering related services to RMA, which are either protocol management services or protocol data transfer service, which we are about to detail here. Each service has a service name in the form of RMA<SERVICE>-<ACTION>, for example RMA-ACTIVATE-Request. Wireless Access in Vehicular Environments ◦ Protocol management services: Protocol management services cover RMA-ACTIVATE-Request, R M A - A C T I VAT E - R e s p o n s e , RMA-NOTIFY-Indication, RMANOTIFY-Confirmation, RMATERMINATESESSION-Indicate, RMATERMINATESESSIONConfirmation, RMA-DEACTIVERequest, and RMA-DEACTIVATEResponse, which ▪ Activate request service: This service is named RMAACTIVATE-Request. Once the RMA is ready to communicate with RCP, it notifies the RMA through the RMA-ACTIVATERequest service provided by RM. ▪ Activate response service: This service is named RMAACTIVATE-Response. When the RM has completed its activation, it gives a response to RMA. ▪ Notify indication service: This service is named RMA-NOTIFYIndication. Once RMA is activated, it would wait for its partner. Through this service, RMA can be informed once the RCP of a OBU has entered the communication zone of the service provider. ▪ Notify confirmation service: This service is named RMANOTIFY-Confirmation. Once the RMA received RMA-NOTIFYIndication, it would respond with a RMA-NOTIFY-Confirmation. ▪ Terminate session indication service: Named RMATERMINATESESSIONIndication. When a transaction is completed, RMA would ◦ send to RM a RMATERMINATESESSION-Indicate. ▪ Terminate session confirmation service: Named RMA-TERMINATESESSIONConfirmation. RM would notify RMA through RMATERMINATESESSIONIndicate that RMATERMINATESESSION-Indicate has been received. ▪ Deactivate request service: Named RMA-DEACTIVERequest. When RMA wishes to stop the operation, it would send this service to notify RM. ▪ Deactivate response service: Named RMA-DEACTIVATEResponse. When RM received R MA -DE A C T IV E -R eq u e s t , it would respond with RMADEACTIVATE-Response. Protocol data transfer services: Protocol data transfer services covered RMA-EXCHANGE-Request, RMAEXCHANGE-Response, and RMAEXCHANGE-Confirmation: ▪ Exchange request service: Named RMA-EXCHANGERequest. RMA uses to send command sequences to RCP. ▪ Exchange response service: Named RMA-EXCHANGEResponse. Once command sequence is sent to RCP and the response sequence is received, RM would reply to RMA with RMAEXCHANGE-Response. ▪ Exchange confirmation service: Named RMA-EXCHANGEConfirmation. When RMA gets the response of RM, this message would be sent back to RM. 105 Wireless Access in Vehicular Environments reFerences IEEE1609.1TM (2006). IEEE Trial-Use Standard for Wireless Access in Vehicular Environments (WAVE) – Resource Manager. IEEE1609.2TM (2006). IEEE Trial-Use Standard for Wireless Access in Vehicular Environments (WAVE) – Security Services for Applications and Management Message. 106 IEEE1609.3TM (2006). IEEE Trial-Use Standard for Wireless Access in Vehicular Environments (WAVE) – Networking Services. IEEE1609.4TM (2006). IEEE Trial-Use Standard for Wireless Access in Vehicular Environments (WAVE) – Multi-channel Operation. Section 3 Location Based Services 108 Chapter 7 Introduction to Global Satellite Positioning System (GPS) Jenq-Muh Hsu National Chiayi University, Chiayi, Taiwan, R.O.C. absTracT Understanding the right positions and directions of people and objects is a significant issue from the ancient eras to the present. In the past, people often launched a war in order to satisfy the craving for the dominating powers and spread their realms. In the recent, Global Satellite Positioning System (GPS) has become the one of most popular positioning technologies. GPS can provide users precise positioning information, no matter wherever that may present their own positions. The early GPS positioning technology has been widely used in military, marine use, until recently gradually applied into our daily life, e.g., automotive navigation, geodesy surveying, etc. In this chapter, we will briefly introduce some GPS issues including the origins of GPS, GPS system architecture, and related GPS applications. inTroducTion Understanding the right positions and directions of people and objects is a significant issue from the ancient eras to the present. In the past, people often launched a war in order to satisfy the craving for the dominating powers and spread their realms. When the army is marching during the fighting, it does not allow the army to get lost in the woods or the dense fog. It needs a method or a tool to offer the army to navigate and move to the right place. DOI: 10.4018/978-1-60566-840-6.ch007 South Pointing Chariot (South Pointing Chariot, 2009) invented in the ancient Chinese civilization is a complex-gearing mechanism without using the magnet to point the same direction while the chariot is around in any movement. South Pointing Chariot only moves on the ground, it can not be used to point the right direction whiling sealing due to the no effect of gearing operation on the sea. The other usual positioning and navigating tool is the compass. The compass has the magnet which can interact with the magnetic field of the earth to always point the same position in north. Although the actual Copyright © 2010, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global is prohibited. Introduction to Global Satellite Positioning System (GPS) positioning is a magnetic North Pole, not a true North Pole, it is enough to use in positioning and navigation. Therefore, the compass has been widely used in many positioning and navigating applications. The compass still has a little accurate positioning problem affected by magnetic interference in natural environment, but it has no serious influence on using the compass. Nowadays, the most famous positioning and navigation service is Global Satellite Positioning System (GPS) (Global Positioning System, 2009; EI-Rabbany, 2002; Kaplan et al., 2006). It uses the satellites orbiting around the earth in space to broadcast positioning signals. A GPS receiver receives the signals and calculates out its current position. That is, GPS provide a worldwide positioning and navigation service for any kind of communication and transportation, such as aircrafts, vessels, vehicles, pedestrians, etc. GPS is developed by the U. S. Department of Defense (DoD) in early 1970s. At the beginning of constructing GPS, it is only used for military purpose, but it has freely opened for civilian use on July 2000s. Therefore, GPS is now a dual-use system that can be accessed by both military and civilian users in various positioning precision for some national security factors considered by U.S. government. As mention above, vehicles having built-in GPS navigation system will be a main trend to provide the driving assistance and navigation service in the future. In next section, we will describe the principle of GPS system in detail. the GPS receivers to receive enough GPS signals without bidirectional communications among them. GPS receives will be able to calculate out the user positions they locate now. GPS system generally consists of a constellation of 24 operational satellites (Leick, 2004). In order to ensure continuous worldwide converge for GPS positioning service, six orbital planes of satellites are organized and each four satellites are placed in an orbital plan. There are four to ten GPS satellites will be visible anywhere in the world under this constellation geometry. The sketch map of GPS constellation is shown in Figure 1. A GPS satellite routes around the earth in a nearby circular orbit, an elliptical shape, with an inclination of 55 degrees to the equatorial plane. The maximal radius of GPS orbit is about 26,560 kilometers measured from the earth center. The orbit period of GPS satellite is approximately 12 sidereal hours, which is about 11 hours and 58 minutes. Thus, GPS satellite will run around the earth twice per day. In order to ensure the availability of GPS positioning service, the number of satellites in the GPS constellation has always been more than 24 operational satellites. Figure 1. The GPS constellation principle oF global posiTioning sysTem The GPS is a satellite-based positioning and navigation system. GPS provides continuous positioning and timing information anywhere in the world under any weather conditions. GPS is also a passively one-way ranging system and it can serve unlimited number of users. That is, users only take 109 Introduction to Global Satellite Positioning System (GPS) The GPS system is comprised of three segments shown in Figure 2: space (satellite constellation), control (ground-control/monitoring network), and user (receiving equipment). The detailed functions of each segment are listed as follow: 1. 2. Space segment: It consists of 24-statellite constellation. It also has some back-up satellites to ensure the full operational capability (FOC) of GPS positioning service. Each satellite transmits a unique PRN (Pseudo Random Noise) ranging signal to measure the distance from GPS receiver to GPS satellite. The satellite signals are controlled by atomic clocks with high precision to transmit through two L-bands, L1 (1,575.42 Hz) and L2 (1,227.6 Hz). GPS satellite has an S-band antenna to communicate with control segment in order to maintain the operational control of GPS service. Control segment: It consists of a master control station (MCS) and a worldwide network of five monitor stations and four Figure 2. GPS segments 110 3. ground antennas shown in Figure 3. The main task of control segment is responsible for tracking, monitoring, commanding, and controlling the GPS satellite constellation in order to determine and predict satellite locations, system integrity, behavior of satellite atomic clocks, and other considerations. The observation of GPS satellites collected at the monitor stations are transmitted to the MCS for processing. MCS sends the processing outcome, the fresh navigation data (control information), to the monitor control stations with ground antenna uploading them to GPS satellites through S-band link. Therefore, the control segment is an important role to maintain the operational behavior of GPS positioning service. User segment: It typically refers to as a GPS receiver shown in Figure 4. When a user uses a GPS receiver to receive enough the L-band signals from GPS satellites under a spacious area, the GPS receiver can determine user’s position and related information, including Introduction to Global Satellite Positioning System (GPS) Figure 3. Geographic distribution of control segment facilities Figure 4. Various types of GPS receivers longitude, latitude, altitude, time, velocity, etc, anywhere in the world. gps posiTioning model The basic idea of GPS positioning model is very simple. It uses the TOA (Time of Arrival) ranging model to measure user position. That is, the distances from a point holding a GPS receiver on the surface of the earth to three GPS satellites are known along with satellite positions, the position of the point can be determined by triangular positioning theory. The concept of TOA ranging model is to measure the propagation time of signal broadcast from signal emitter at known locations. Assume that the signal emitter regularly broadcast the beacon signal and both of signal emitter and 111 Introduction to Global Satellite Positioning System (GPS) receiver have synchronous clocks. The receiver can measure the propagation time of beacon signal from emitter and then propagation time is multiplied by the speed of the signal to obtain the distance from the emitter to the receiver. GPS system provides three-dimensional position information including longitude, latitude, and altitude. It needs to use the three-dimensional position determination to calculate the position of the measured position. In order to simply examine the GPS positioning model, we first introduce the two-dimensional position model how it achieve the point positioning. Consider a mariner at sea determine his vessel’s position from a foghorn. Assume that the vessel and the foghorn are equipped with accurate synchronized clocks, the foghorn whistle is repeatedly sounded per minute, and the position of foghorn is known. The marine measures the elapsed time from minute mark until the foghorn whistle is heard. Thus, propagation time (t) of foghorn whistle multiplies speed (v) of the sound is the distance (R, R = v × t) from the foghorn to the marine. Let the foghorn be denoted F1 and the distance from the foghorn F1 to the marine be denoted R1 respectively. Thus, marine only knows the vessel is somewhere on a circle with radius R1 centered about the foghorn, which is shown in Figure 5-(1). If there is the second foghorn F2 and the marine can simultaneously measures the distance R2, the vessel position will be located at one of the intersections of the range circles with radius R1 and R2 centered about the Figure 5. Two-Dimensional Position Determination: (1) All possible positions ranging from a single source, (2) Two ambiguous positions from measurements to two sources, and (3) Position ambiguity removal by three sources. 112 Introduction to Global Satellite Positioning System (GPS) foghorns F1 and F2 respectively, which is shown in Figure 5-(2). In order to resolve the positioning ambiguity of the vessel, third foghorn F3 is applied to measure the distance R3 for removing position ambiguity as shown in Figure 5-(3). In fact, the TOA measurement would not be perfect due to the clock drift, sound interference, atmospheric effects, etc. These errors would affect the measurement and resulting in accurate distance computations. GPS also employs TOA ranging and threedimensional positioning model to calculate user position. The measurement of GPS positioning model is similar to two-dimensional positioning model. Assume that each satellite transmits a unique ranging signal in space. A satellite has a clocks synchronized with others clocks built on other satellites. The clock synchronization and orbit plane of these satellites are monitored and controlled by the control segment (MCS). A GPS receiver also has a clock synchronized to GPS system time. Ranging signals transmitted from GPS satellites contain time information. It enables the receiver to measure the propagation time (t1) of the ranging signal left the satellite (S1) based on satellite clock time. Thus, the satellite-to-user distance (R1, R1 = Vlightspeed × t1) can be calculated. The user would be located somewhere on the surface of a sphere centered about S1, as shown in Figure 6-(1), in this ranging measurement. If the second satellite (S2) is considered into account, the user would be located somewhere on the surface of both spheres ranging from S1 and S2, as shown in Figure 6-(2). That is, the user position is located at a plane of intersection. The third satellite (S3) is adopted to measure the user position. The user would be located at the interaction of the perimeter of the circle and the surface of the sphere centered about S3, as shown in Figure 6-(3). We can find that the user position would be at two possible points, and only one of the points is the correct user position. For the user position, it is generally locates on the surface of the earth. Therefore, the lower one of two points will be the true position. In Figure 6. GPS Three-Dimensional Positioning Model: (1) User located on surface of sphere, (2) User located on perimeter of plane of intersection, and (3) User located at one of two points. order to measure the more accurate user position, the fourth satellite will be applied to calculate and fix the user’s altitude. In addition GPS can determine the user velocity. The most popular used method is based on estimating the Doppler frequency of the received GPS signals. GPS also can measure the moving orientation (direction) of a rigid object, e.g., a vessel, an aircraft, etc. gps reFerence coordinaTe sysTem Although GPS can measure the user position anywhere, under any weather conditions. Due to the irregular topographic surface of the earth, the determination of the user position is very difficult. To over this problem, it needs a mathematical formulation to model the smooth surface of the 113 Introduction to Global Satellite Positioning System (GPS) earth. Therefore, it needs a reference coordinate system to represent the states of GPS satellites and the receiver. A coordinate system is a set of rules to specify the locations of points. This usually specifies the origin (the central) and a set of reference lines (the axes) with known orientation. Figure 7 shows a 3-D Cartesian coordinate system which uses three reference axes, x, y, and z, intersecting at the origin (C) of the coordinate system. Coordinate systems can be classified according to the reference surface, the origin, and the orientation of the axes. For example, in a 3-D geodetic coordinate system (Torge, 1991), the reference surface is an ellipsoid. The orientation of the axes and the origin are specified by two planes: the meridian plane through the polar and equatorial plane. It is important to provide an easy-understanding positioning information for the GPS users. Thus, positioning information in a 3-D geodetic coordinate system can be described in the geodetic latitude (ϕ), the geodetic longitude Figure 7. A sample of 3-D coordinate system 114 (λ), and the altitude (h) above the reference surface. That is, geodetic coordinate (ϕ, λ, h) of Point P can be easily transformed to or from Cartesian coordinate (x, y, z). The world geodetic system (WGS) (World Geodetic System, 2009) is a standard coordinate system used in cartography, geodesy, and navigation. It consists of a standard coordinate frame for the earth, a standard spherical reference surface for raw altitude data, and a surface (nominal sea level) with the same gravitation. That is, The WGS can provide an ellipsoidal model to represent the real earth shape. Therefore, the WGS can provide the means for relating positions on various local geodetic systems to an Earth-centered, Earth-fixed (ECEF) coordinate system. A series of WGS revolution, WGS 60 (developed in 1960), WGS 66, WGS 72, and WGS 84, have been developed to provide more accurate coordinate system. Currently, the GPS system uses the WGS 84 to be the standard of coordinate reference system. Introduction to Global Satellite Positioning System (GPS) gps daTa proTocol FormaT Different GPS manufacturers may have their own data formats to store the GPS measuring information. It may lead to increase the difficulty of system integration. Thus, it is necessary to define the standards of GPS data formats to access the uniform GPS information from various GPS receivers. A number of standard formats of GPS information representation for various needs have been developed, which includes RINET, NGS-SP3, RTCM SC-104, NMEA 0183, etc. NMEA 0183 is the most popular GPS data format in these standards. The National Marine Electronics Association (NEMA) has developed a specification defining the standard, called NMEA 1083, (NMEA Data, 2009) communicating interface among various marine electronic equipments and application software. The standard allows marine electronics to send information to computer or to other marine equipment. GPS receiver communication is also defined within the NMEA standard. GPS receivers generally use the RS232 protocol (EIA-422) to communicate with GPS software through computer serial ports. The interface speed of NMEA standard is 4800 baud rate with 8 bits of data, no parity, and one stop bit. Thus, all GPS receivers should support this speed. But some modern GPS receivers can support higher interface speeds, e.g., 9600 or 38400 bits per second, to communicate. Data stream in the NMEA 0183 standard is an ASCII format. It includes the complete position, velocity, and time (PVT) information. The data is sent in the form of sentences. Each sentence begins with a ‘$’ character and ends with a carriage return/line feed <CR><LF> sequence. The maximal length of sentence is less than 80 characters of visible text. The data fields in sentence formed in a single line are separated by commas. The first 5 letter prefix following the beginning character ‘$’ identify the sentence type (first 2 letters) and sentence content (successive 3 letters) in a sentence. The prefix of sentence type is GP for GPS receivers. The last field in any sentence is a checksum field following a delimiter character ‘*’. NMEA standards have defined various GPS sentences. Table 1 shows some common GPS sentences used in a GPS receiver and an example for decoding GPGGA sentence is illustrated in Table 2. Besides, we also can visit the web site (http:// gpsinformation.org/dale/nmea.htm) to view the detailed structures and formats of other common GPS sentences. Figure 8. An example of valid GPS sentences generated from a GPS receiver 115 Introduction to Global Satellite Positioning System (GPS) Table 1. Some common GPS sentences used in GPS receiver Name GPS information can be visualized displayed on the GPS viewer software or equipment. Figure 9 depicts an example of GPS viewer showing GPS information including the longitude, latitude, height, speed, direction, the number of satellites in view of the sky, their locations of viewed satellites, and the strength of GPS signals. Notes GPGGA Global positioning system fixed data GPGSA GPS DOP and active satellites GPGSV Satellites in view shows data about the status of satellites GPRMC Recommended minimum command GPVTG Velocity information GPDLL Geographical Latitude and Longitude gsp applicaTion For Vehicle naVigaTion Since the price of GPS receiver has become cheaper, the usage of GPS has also become more ubiquitous. Thus, many innovative GPS services including navigation and positioning, tracking, and surveying, etc., have been widely applied into various application domains. In this section, a GPS Figure 8 shows a series of GPS sentences generated from a GPS receiver, which sentences contain five main content types: GPGGA, GPGSA, GPGSV, GPRMC, and GPVTG. After decoding a series of GPS sentences, which include GPGGA, GPGSA, GPGSV, GPRMC, and GPVTG, these Table 2. The decode of GPGGA sentence GPGGA Sentence Data Items in Sentence $GPGGA,040905.000,2333.2209,N,12025.6358,E,1,07,1.2,37.6,M,17.1,M,,0000*61<CR><LF> $GPGGA,[1],[2],[3],[4],[5],[6],[7],[8],[9],[10],[11],[12],[13],[14],*[15]<CR><LF> Number Example [1] Time of position in UTC system [hhmmss.ss] 040905.000 [2] Latitude [lll.ll] 2333.2209 [3] North or South [N/S] N [4] Longitude [yyyy.yy] 12025.6358 [5] East or West [E/W] E [6] GPS quality indicator [0=invalid; 1=GPS fix; 2=Diff. GPS fix, …] 1 [7] Number of satellites in use [xx] 07 [8] Horizontal dilution of position (HDOP) [x.x] 1.2 [9] Antenna altitude above/below mean sea level (geoid) [x.x] 37.6 [10] Meters (Antenna height unit) [M] [11] Geoidal separation (Diff. between WGS-84 earth ellipsoid and mean sea level. -=geoid is below WGS-84 ellipsoid) [x.x] [12] Meters (Units of geoidal separation) [M] [13] Age in seconds since last update from diff. reference station [x.x] [14] Diff. reference station ID# [0000-1023] * [15] <CR><LF> 116 Notes Checksum delimiter character [*] Checksum (hh) Sentence terminator [<CR><LF>] M 17.1 M (NULL) 0000 * 61 <CR><LF> Introduction to Global Satellite Positioning System (GPS) Fgiure 9. An example of GPS viewer virtualizedly showing GPS information application concentrating on vehicle navigation will be discussed. GPS provides the location, velocity, orientation, and time information that enables many practical applications to be used in our daily lives. The first popular application is GPS navigation service (Drane et al., 1998). A car installed an in- vehicle GPS navigation system (automotive GPS navigation system) can assist the driver to follow a route plan to reach the driving destination s/he sets. Figure 10 shows an example of using GPS navigation to be an aid for driving during a trip. Although GPS navigation system can acquire the position information from GPS satellite signals, it needs a digital map stored in the system or retrieved from the Internet to locate the user or the car on the road in a road network. Thus, a newest full set of digital map is a very significant in order to provide an accurate GPS navigation for matching to the real road network environment, e.g., a new road has been opened. A friendly userinterface for user is also an important issue in a GPS navigation system. A modern GPS navigation system may consist of visual-display interface (e.g., virtual road network in 3D view), touch screen (e.g., target input), voice-based interface (e.g., speech recognition and synthesis), etc., to satisfy the various requirements of the GPS driving guidance. Some GPS navigation systems may have a module, to receive the real-time traffic information to dynamically update the navigating guidance service fitting the real road network condition via traffic message channel (TMC) (Traffic Message Figure 10. Use the GPS navigation system to aid the drive 117 Introduction to Global Satellite Positioning System (GPS) Channel, 2009) through FM radio broadcasting, WiFi/GPSR/3G data transmission, inter-vehicle communication, etc. In addition to GPS navigation service, the GPS can also be used in vehicle tracking, GPS security, fleet management, intelligent public transportation service, emergency rescue, biological movement, geodetic surveying, graphic information system, location-based service, etc. We also find that a cellular phone with a built-in GPS receiving module is a trend. It means that cellular phone will act a portable and personal GPS location and navigation platform to provide the walker navigation with highly accurate guiding information, the location-aware services of the four essential requirements of the people: food, clothing, housing and transportation, in the future. conclusion In this chapter we have briefly introduced an essential principle of the GPS system, including GPS constellation, GPS segments, GPS positioning model, standard GPS data format, and GPS application for vehicle navigation. There are still many issues, such as GPS satellite signal characteristics and acquisition, GPS signal inference and multipath, error measurement, differential GPS (D-GPS) (Differential GPS, 2009), assisted GPS(A-GPS) (Assisted GPS, 2009), other GPS systems, geodetic/geocentric coordinate system, map projection, etc., to be discussed. reFerences Assisted, G. P. S. (n.d.). Retrieved September 11, 2009, from http://en.wikipedia.org/wiki/Assisted_GPS. 118 Differential, G. P. S. (n.d.). Retrieved September 11, 2009, from http://en.wikipedia.org/wiki/Differential_GPS. Drane, C., & Rizos, C. (1998). Positioning system in intelligent transport systems. Boston: Artech House. EI-Rabbany. A. (2002). Introduction to GPS – The Global Positioning System. Boston: Artech House. Global Positioning System. (n.d.). Retrieved September 11, 2009, from http://en.wikipedia. org/wiki/Global_Positioning_System. Kaplan, E. D., & Hegarty, C. J. (2006). Understanding GPS – Principle and Applications. Boston: Artech House. Leick, A. (2004). GPS Satellite Surveying. New York: John Wiley & Sons. NMEA Data. (n.d.). Retrieved September 11, 2009, from http://gpsinformation.org/dale/nmea.htm. South Pointing Chariot. (n.d.). Retrieved September 11, 2009, from http://en.wikipedia.org/wiki/ South_Pointing_Chariot. Torge, W. (1991). Geodesy. Berlin: Walter de Gruyter. Traffic Message Channel. (n.d.). Retrieved September 11, 2009, from http://en.wikipedia.org/ wiki/Traffic_Message_Channel. World Geodetic System. (n.d.). Retrieved September 11, 2009, from http://en.wikipedia.org/wiki/ World_Geodetic_System Zhao, Y. (1997). Vehicle Location and Navigation Systems. Boston: Artech House. 119 Chapter 8 Vehicle Location and Navigation Systems Ben-Jye Chang National Yunlin University of Science and Technology, Yunlin, Taiwan, R.O.C. absTracT The most driving purpose is to traverse to the destination safely, efficiently, and comfortably. Two types of approaches could achieve the goals, including the static and dynamic approaches. In the static aspect, vehicles use the static road and traffic information to navigate. Conversely, in the dynamic aspect, vehicles adopt the dynamic information instead. However, both of the two approaches first require getting the vehicle’s location and then map the position on an e-map. Thus, this chapter first introduces some important vehicle location determination algorithms: the dead reckoning and global position system algorithms, in which the precision of location technologies are compared. Then, the map-matching algorithm is described in detail. Finally, various vehicle navigation approaches are detailed, in which the important topics include: the navigation architecture, the navigation routing algorithm, and navigation applications. inTroducTion The invention of the vehicles, though it could shorten the distance from place to place, has brought a lot of serious issues: traffic jam, car accident, energy crisis, environmental pollution, etc. Moreover, as the population rises and living in standard increases constantly, the vehicles are everywhere and the roads are even more complicated. Finally, it would not DOI: 10.4018/978-1-60566-840-6.ch008 only waste our time, but social cost. Therefore, transportation simplification, driving time reduction, and energy conservation are the top priority concern of the world. Thus, this chapter, Vehicle Location and Navigation Systems (Farrell et al., 2008; Zhao et al., 1997), are what we researched. While the vehicles are moving, realizing your location, destination and the shortest path are important. However, most of all, you know where you are, or other conditions are useless. The early days of vehicle position are a driver reads maps and road signs in the meantime Copyright © 2010, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global is prohibited. Vehicle Location and Navigation Systems and finally reaches its destination. It is inefficient and dangerous still. Moreover, wasting resources and other problems could occur by vehicles position. Generally, vehicle position can be divided into two methods: early days’ dead position and current used GPS technologies. For locating a vehicle position, the vehicle with dead reckoning method must set its coordinates before moving. Position and distance sensors, based on the coordinates, could figure out tracking movement and turning degree. Lastly, the current position could be located. For example, as shown in Figure 1, the starting point (x 0 , y 0 ) , the based point also, multiplied by each distance di and aspect qi is the terminal point (x n , yn ) . The formula becomes as follows: n x n = x 0 + å diT cosqi i =0 n yn = y 0 + å diT sinqi (1) i =0 Distance and position sensors are applied by dead reckoning method. The following contents in this book would discuss both two common sensors. Distance sensor normally applies speedometer to plot the moving distance. The application makes use of the sensor device, which can calculate wheel turns, on the wheel axle. Circumference of the wheel multiplied by the wheel turns is the distance. Figure 1. Dead reckoning Figure 2. Distance sensor device However, the results could be a huge error due to different wheel sizes and terrains (e.g., idle running of wheels on the snow and sand). Distance sensors are subdivided into mechanical and electronic ones. Figure 2 is a typical example. The vehicle position sensors enable position measurement. It can either be an absolute position sensor or a relative one. The first one applies the earth’s magnetic field to get absolute position. Figure 3 shows that the induction coils can be affected by the earth’s magnetic field and the vehicle’s position can be measured can by the angle q , as demonstrated in equation 2. However, using earth magnetic field to induct potion has a disadvantage. That are the results are easily affected by some circumstances: viaducts, bridges, substation, and high-voltage tower. Conversely, the relative position is determined in terms of the original point and angle. q = tan-1 Vx Vy (2) Dead reckoning was mainly used in the past with three disadvantages: 1) Based point should be located well, 2) The sensor has huge error, and 3) The more distance you move, the more error could be. 120 Vehicle Location and Navigation Systems Figure 3. Earth magnetic sensor global positioning system Since the first generation GPS chip was presented to the public, it overcomes some problems: unclear signal of ex-chip, long positioning time, location drifting, etc. Due to the progress of chip manufacturing procedure and size reduction, even PDA, digital camera, cell phone and so on can be integrated with GPS. The GPS device equipped on the vehicles is simple and more accuracy than dead reckoning method. It is getting to be used as basic vehicle equipment. Conversely, some disadvantages of GPS still need to be addressed. For example, if a cell phone or radio’s wave band is closed to GPS’, it would cause interference. In addition, tunnels and weather would affect the accuracy of GPS also. Thus, adjusting GPS error by a fixed base station and then conveying it to the vehicles’ GPS for correcting can add accuracy. It is called Differential GPS (namely D-GPS). As Figure 4 demonstrated, the FM radio can also receive the positioning signal from the base station. Figure 4. Differential GPS (D-GPS) 121 Vehicle Location and Navigation Systems map maTching The vehicle positioning system is different from the airplane and ship positioning systems. The reason is that the vehicle movement tracks must on roads. Recently, maps have already been digitized; thus, the error between digital maps and movement tracks can be corrected by computers. As a result, a more accurate position can be resulted, as shown in Figure 5. map matching method The theorem of map matching is to match a vehicle moving track on a digital map (Brakatsoulas et Figure 5. Map matching Figure 6. Digital network map 122 al., 2005). In a digital map, a road is viewed as a link and an intersection is viewed as a node, as shown in Figure 6. Furthermore, when a vehicle is moving, the coordinates of the vehicle can be compared with the digital map, as demonstrated in Figure 7. For example, the black bold line, namely a road, and the green one is the vehicle movement track, in which we could get two vertical lines: L1 and L2. If L1 is shorter than L2, it means the vehicle movement track is on road A, and vice versa. This is the basis theorem of the map matching. Although the vehicle positioning can use map matching to correct the movement tracks and to get accurate position, map matching still has some Vehicle Location and Navigation Systems Figure 7. Map matching drawbacks. For example, if a road is just not on the digital map or one drives to parking lots or open ground, how to calculate position with map matching? Therefore, an error upper and lower range should be set. Once an error is in the range, it initiates the correction of the movement tracks; otherwise, the error can be ignored. Vehicle Tracking The GPS device is easy to be carried and easy to be integrated with other systems. Moreover, due to the common used and free charge of using GPS, the vehicle position can be opened another different applications, e.g., the vehicle tracking, the transportation online information, etc. introduction of Vehicle Tracking The vehicle tracking issue is a vehicle equipped with GPS and has the capability to report its coordinates to a computer periodically. In Figure 8, a user can trace the target vehicle’s position through the Internet. Similarly, more and more new applications of vehicle tracking are being created, for example, a remote alarm system, transportation tracking, etc. Figure 8. Vehicles tracking 123 Vehicle Location and Navigation Systems model of the Vehicle Tracking system Basically, the model of the vehicle tracking system can be divided into two parts: the position dispatch end and the reception end. In the position dispatch end, the GPS receiver continuously receives the satellite signal and then locates the vehicle’s position. Meanwhile, the coordinates can be dispatched to the Internet through the wireless network base stations. The position reception end can receive and match the coordinates from the Internet and pass current position to the digital map. Finally, the date can be shown on the screen, as demonstrated in Figure 9. The applications of the Vehicles Tracking The applications of the vehicle tracking are very extensive. It is even more than what we can think about of. Till now new applications are still brought up. However, normally they can be divided into several parts. Security: • Cash-carrying vehicle security surveillance: So far the vehicles tracking system, a basic equipment for a cash-carrying vehicle, can track a vehicle’s position Figure 9. Fame of the vehicles tracking system 124 • effectively to avoid embezzlement and be managed easily. Car alarm system: The best car alarm system equipped also by car rental agencies now is the vehicles tracking system. Once the vehicle was stolen, we can find out no matter wherever it is. The distribution and management system of police cars and ambulances through surveillance center management. If something urgently happened, they can send the nearest police cars or ambulances to support. Furthermore, the patrolling position and the public security can be improved efficiently. Transportation: • Cargo services: Through the vehicles tracking system, we can know cargo destination, map out a route to avoid wasting time. A consumer can realize the cargo’s status through the Internet. The delivery of dangerous goods and polluted material, e.g., through the vehicle tracking system, once chemical or toxic material, stone delivery and the like are leaking out, we can deal with it in the very beginning and avoid other vehicles entering the limited area. Vehicle Location and Navigation Systems Public transit: • • Vehicles distribution: Before you head somewhere, through the Internet, one can check the nearest taxi or bus current position. Moreover, through the vehicle tracking system, speeding and route time can be estimated. The management of a fleet of vehicles: It is not only for making up for cell phone using defects, but also avoiding lagging behind and getting lost. auTonomous locaTion and naVigaTion In the past, the cost of navigation system was very high, so that it was used on special applications. In increase of time, the cost is down and drivers use the navigation system to locate the position and to navigate without any help. In this section, we briefly depict the issues of the autonomous location and the navigation systems. autonomous location The autonomous location is that a vehicle can locate the position by itself. In (Zhao, 1997; Zhao et al., 1997), authors classified the location technologies to three types: the stand-along location technology (Chang et al., 2008), the terrestrial radio technology and the satellite technology. In the stand-alone location technology, the Dead Reckoning is an example to use the self ability to locate the position. The terrestrial radio technology and the satellite technology are similar, because the terrestrial radio technology adopts the signals of base stations to locate the position and the satellite technology uses the satellite signal to locate position. The precisions of different technologies are indicated in Table 1 (Schiller et al., 2004). Based on the results, only using one location technology is not enough for some applications, because the Table 1. The precisions of different location technologies Technology Medium Precision GPS Radio 25 m DGPS Radio 3m WAAS Radio 3m MPS Radio 150 m coordinates computed by a location technology may exist significant erroneous. By integrating with several location technologies is a good approach for reducing errors. autonomous navigation Navigation means that a position can be located on a map and a routing algorithm referring to the location of the vehicle calculates the path for a driver. Autonomous navigation means that navigation is done by self. Figure 10 proposes an example of the autonomous navigation architecture (Zhao, 1997). The communication devices are used to receive the radio signals and to calculate the position. After determining the position, the user can input the destination for the system to determine the routing path. Figure 10. The architecture of a navigation system 125 Vehicle Location and Navigation Systems A navigation system not only provides the routing function, but also offers additional functions such as weather bulletins, restaurant position, etc. Based on the factors, designing a navigation system must consider what typical kinds of functions should be included in the navigation system. cenTraliZed locaTion and naVigaTion After describing the navigation and location technologies in the previous section, this section details the location and navigation systems based on a centralized framework. In the centralized location and navigation systems, all functions are done in a centralized server for reducing the computing loading of clients. In this section, we first introduce the architecture of centralized systems. Location architecture and navigation architecture are then depicted. centralized architecture Zhao (1997) shows the Centralized architecture, as shown in Figure 11. Communication devices are adopted to communicate with the wireless technology such as WiMAX and GSM in order to perform the communication between the client and the server. A client requests a service from server and the server provides a service to the client. location architecture In Figure 12, Zhao (1997) introduced a location architecture of a centralized system. This architecture consists of three parts including communication devices, a client and a server. The client uses the GPS device or other methods to get the coordinates, and then sends it to the server. When receiving the coordinates of the client, the server computes the position and locates it on a digital map. The final result is shown on a graph-based interface, as demonstrated in Figure 13. Figure 11. System relation between communication devices, client and server Figure 12. Location architecture of a centralized system 126 Vehicle Location and Navigation Systems navigation architecture In Figure 14, Zhao (1997) introduces a navigation architecture of a centralized system. In the client, after keying the destination, the client sends a navigation request to the server. When the server receiving a request, the server determines a path for the client and then sends the determined routing path to the client. According to the received results, the client drives to the destination via the central-computed routing path. auTomaTic Vehicle driVing and naVigaTion Automatic vehicle driving technology let user drives the vehicle without any operations. How to achieve this goal? In this section, we thus discuss these technologies that automatic vehicle needs. Firstly, we introduce automatic vehicle driving and navigation. The needs of technologies of automatic vehicle are then discussed in detail. introduction In the past, an automatic vehicle was a dream but now technologies are advanced and can be accomplished. The traffic accidents are often happened when drivers are scatterbrained or other factors which are not avoided. By using the automatic vehicle, it could reduce the probability of accident. An automatic vehicle also saves us a lot of time which we could use to do other things. For realizing an automatic vehicle that can automatically drive to a destination, some key technologies such as image process, navigation system and automatic control system are required. Figure 13. (a) A scenario of simulation (b) The display result of location Figure 14. Navigation architecture of a centralized system 127 Vehicle Location and Navigation Systems These technologies need a powerful computer but may have some problems. In the increase of time and advance of technologies, these problems will be overcome and vehicles will have a function of automatic driving. consist of automatic Vehicle From the descriptions of previous sections, we know that the automatic vehicle will be the trend in the future. An automatic vehicle consists of a computer vision system, a navigation system and a vehicle control system. The relation of these systems is demonstrated in Figure 15. First, the computer vision processes a series of the image, which is capture by a camera, and then returns the process result to the vehicle control system. Second, the vehicle control system controls the vehicle according to receive an information from other systems. It only has the computer vision and vehicle control systems are not enough because the vehicle is lack of an indicator for the use of arbitrarily drive. Based on this condition, the provision of a navigation system for an automatic vehicle is necessary and important. The navigation system and the computer vision system are related with each other, because if there is an error of location in the navigation system, the computer vision system can be used to correct Figure 15. The components of automatic vehicle Figure 16. Computer vision system 128 the vehicle location. For example, the navigation system instructs the vehicle to turn right in 150m, but it must turn right in 100m. In order to reduce the error condition, we use a computer vision system to correct the error. The combination of a computer vision system, an automatic vehicle control system and a navigation system has the capability to achieve an automatic vehicle. Navigation System In previous sections, the navigation system consists of a location system and map information system which was introduced in Chapter 13. The navigation system plays a very import role in the automatic vehicle that instructs the vehicle when the vehicle makes turns. If a vehicle is lack of the navigation system, it will drive randomly and causes it dangerously. That is, a navigation system is an instructor telling a location with a small error. The development problem is how to design a system with high precise location and how to use other related methods to improve the decision of automatic driving. Computer Vision System A computer vision system consisting of a camera and image process observes environments and Vehicle Location and Navigation Systems responds to the control system. Figure 16 shows an example of a scenario. Firstly, the camera captures images converted to digital images, and then the image process detects some targets. In consequence, the result of detection is shown in the original image. The function of a computer vision system can be divided into three types: the lane detection, the vehicle detection and the obstacle detection. First, the lane detection helps an automatic vehicle to detect a lane which the vehicle drives on. Second, there are many vehicles driving on a lane. If a system can’t detect other vehicles, it will cause accidents. Based on this reason, vehicle detection is necessary. Finally, a lot of objects may be on lanes such as people, animals, rocks, traffic light, etc. The obstacle detection helps the automatic vehicle to detect these situations and to respond to the control system immediately. In this subsection, we first introduce the lane and vehicle detections. In the lane detection, Jung et al., (2005) introduced a lane detection technology. The procedures include edge detection and boundary detection. In the edge detection, the edge distribution function and Hough transform are adopted to detect edges. In the boundary detection, a linear model is adopted to detect boundaries. Figure 17(a) is an original image. The image shown in Figure 17(b) is grayed and then uses boundary detection to find the rough sketch of lanes. Finally, the lanes are drawn in the original image as demonstrated in Figure 17(c). In most related studies, they can detect the marked lanes; however, in the real roads not all of Figure 18. (a) An original image. (b) A grayed image. (c) Vehicle detection. (d) A result of detection roads are marked for each lane. For this reason, how to handle this exception must be addressed. In the vehicle detection, the function of the vehicle detection is to detect other vehicles in front of the detecting vehicle. The vehicle detection must consider some impact factors such as the distance between vehicles, the speed of vehicle, etc. In this subsection, we only consider the vehicle detection without other factors. Broggi et al., (1999) introduced the vehicle detection as indicated in Figure 18. First, it grays the image and then uses the symmetry detection that detects a symmetry part of a vehicle for detecting vehicles. The symmetry detection uses the edge detection technology, and then uses the results of vertical and horizontal edge detections to find an outline of a vehicle. It then uses the bounding box detection, detecting the left and right box on the top or foot, to increase the result of detection. If a Figure 17. 129 Vehicle Location and Navigation Systems Figure 19. Structure of the vehicle control reFerences Brakatsoulas, S., Pfoser, D., Salas, R., & Wenk, C. (2005). On Map-Matching Vehicle Tracking Data. In The 31st VLDB Conference, (pp. 853 – 864). Broggi, A., Bertozzi, M., Fascioli, A., Guarino, C., Bianco, L., & Piazzi, A. (1999). The Argo Autonomous Vehicle’s Vision And Control Systems. International Journal of Intelligent Control and Systems, 3, 409–441. box is not fined, it will use backtracking to correct results and repeatedly uses boundary detection to find vehicles. Vehicle Control System The vehicle control system, an important component of the automatic vehicle, receives information from the computer vision and navigation systems, and then controls the vehicle. If the control system design has defects, it will cause accidents. Based on the reason, providing a stable and robust control system is required. Thus, a control system must consider many impact factors: error correction, speed of driving, etc. However, we only introduce a basic architecture proposed in (Broggi et al., 1999), in which the structure of the vehicle control system is shown in Figure 19. A computer vision system sends a result of image process to the vehicle control instructor, controlling the vehicle according to vehicle information, image process and location. Finally, the vehicle obeys instructor’s command to drive. 130 Chang, B.-J., Huang, B.-J., & Liang, Y.-H. (2008). Wireless Sensor Network-based Adaptive Vehicle Navigation in Multihop-Relay WiMAX Networks. The 22nd International Conference on Advanced Information Networking and Applications, (pp. 56-63). Farrell, J. A. (2008). Aided Navigation: GPS With High Rate Sensors (pp. 22-80). New York: Glencoe/McGraw-Hill School Pub Co. Jung, C. R., & Kelber, C. R. (2005). Lane following and lane departure using a linear-parabolic model. Image and Vision Computing, 23, 1192–1202. doi:10.1016/j.imavis.2005.07.018 Schiller, J., & Voisard, A. (2004). Location-Based Services. San Francisco: Morgan Kaufmann. Zhao, Y. (1997). Vehicle Location and Navigation Systems. Boston: Artech House Publishers. Zhao, Y., & House, A. (1997). Vehicle Location and Navigation Systems: Intelligent Transportation Systems (pp. 43-78). Boston: Artech House Publishers. 131 Chapter 9 Design and Implementation of Vehicle Navigation Systems Min-Xiou Chen National Dong-Hwa University, Hualien, Taiwan, R.O.C. absTracT Vehicle Navigation System (VNS) is a complicated and integrated system. A reliable vehicle navigation system should integrate the wireless communication technologies, positioning technologies, embedded computer, geographic information database, and so on. The major purpose of the chapter is to help understanding the architecture of vehicle navigation system. This chapter first introduces the system requirements and system analysis, and show the system platform of vehicle navigation system. The system platform can be divided into six components. There are the digital map database, positioning devices, map-matching process, route planning process, route guidance process, human-machine interface, and wireless communication interface. The design issues and system communication of these components are detail illustrated in the chapter. Finally, the authors also present some vehicle navigation systems proposed in the past few years, and show the difference of these systems. The aim of vehicle navigation system is to guide the vehicle along the optimal path from the starting point to destination. A reliable vehicle navigation system can reduce the traffic chaos in the city and improve the transportation delay. In order to achieve reliable vehicle navigation system, the detail system requirements, system analysis, and system architecture are shown in the chapter. Each component of vehicle navigation system is briefly illustrated, and the system communication is also described. The authors also present the architecture of the proposed vehicle navigation system, and show the difference of these systems. Therefore this chapter helps understanding the architecture of vehicle navigation system. DOI: 10.4018/978-1-60566-840-6.ch009 Copyright © 2010, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global is prohibited. Design and Implementation of Vehicle Navigation Systems inTroducTion system requirements The aim of navigation system is to detect the position of the vehicle, track the vehicle and control the movement of the vehicle from one place to another. The location techniques and the map are two important components in the navigation system. In Age of Exploration, the shipman records their position on the chart and pilots their courses according the compass, sextant, and chronometer. After World War II, the radar technique is implemented to identify the range, altitude, direction, or speed of both moving and fixed objects. The radar technique is also involved in the modem navigation system in order to improve the correctness of the vehicle position, and sailing safety. In recent years, with the development of global positioning technology, electronic technology, and wireless communication technology, the size of navigation system becomes smaller and can be carried on a bus, truck or car. Thus, the automotive navigation system had been proposed to guide vehicles in its location using digital map display. Moreover, with rapid increment of automotives, the urban traffic becomes much more crowed and the traffic chaos is a serious problem in many cities. Therefore, how to design a vehicle navigation system in order to reduce the traffic chaos and improve the transportation delay is a critical problem. In the chapter, the detail architecture of vehicle navigation system is illustrated. Let’s consider a scenario that a visitor is driving in a strange city, how can he know his position and plan the driving path to the targeted place based on his tourist map? Getting lost is a nightmare for each visitor. Thus, the system requirements for the tourist and the strange driver are shown as below (Zhao, 1997): The design oF Vehicle naVigaTion sysTems k. The development process of an information system should involve the system requirements, system analysis, system design, system construction, system testing, and system maintenance tasks. This chapter will focus on the system analysis, system requirements, and system design of the vehicle navigation system. 132 a. b. c. d. e. f. g. h. i. j. l. m. n. Shows the correct position of the current vehicle or destination on the digital map; Plans the shortest/fastest route from the current position to the destination and show on the digital map; Guides the drivers along the planned route; Tracks the vehicle on the digital map; Easy and safety to use; Shows the real-time traffic information (emergency or congestion) on the monitor; Re-plan routers based on emergency information or congestion information. Moreover, more system requirements shown as below had been proposed in the investigation results for the Taiwanese drivers . Shows the warning message of the speed traps or the over speed limit; Shows the park area and the gas station on the digital map; Shows the shopping/restaurant information on the digital map; Shows the tourist information on the digital map; Shows the Road-Side service; Reports the traffic accident; Summarize these system requirements, the basic system requirements can be shown as the following: The function to show the location of the current vehicle or destination on the digital map; Design and Implementation of Vehicle Navigation Systems o. p. q. The functions to plan or re-plan the shortest/ fastest route; The functions to show the life information; Friendly user interface. e. system analysis Analyzing the user requirements, the functions of the vehicle navigation system can be illustrated as following: a. b. c. d. Needs a digital map library: The digital map database is the key component for the vehicle navigation system. The road map should be digitalized into the digital library, before the system uses it. The location information of some government institutions and some famous buildings should be digitalized into the digital library, which can be denoted as the landmarks. The system can obtain and show the relative position of each targeted object on the map according to the location information of these landmarks. Needs the location modules: The location modules can be divided into the positioning module and map-matching module. The major function of the positioning module is to get the position information from the position devices and obtain the coordinates of a vehicle on the surface of the Earth, or the relative position of a vehicle on the city. The aim of the map-matching module it to provide drivers the correctly location information of the vehicle and show the precisely position of the vehicle on the map. Needs a route planning module: The route planning module is a process that provides the drivers to plan the route from the starting point to the destination. Needs a route guidance module: The route guidance module is a process that guides the drives along the planned route from f. the current position of the vehicle to the destination. Needs an interactive interface: The interactive interface involves the input devices and output devices. The drivers can enter the information into the system, or select the information from the system by the input devices, and see the results from the output devices. Needs a communication module: The communication module can receive the current traffic information, such as emergency information and congestion information, from the traffic control center. The system can re-plan a new route based on that current traffic information. The vehicle navigation system also can report the traffic information to the traffic control center. The implemenTaTion oF Vehicle naVigaTion sysTems According to the system analysis, the system architecture of the vehicle navigation system can be shown as the Figure 1. The vehicle navigation system involves six components. There are the digital map database, positioning devices, mapmatching process, route planning process, route guidance process, human-machine interface, and wireless communication interface. The implementation issues of these components are illustrated as following section. system platform a. Digital map database: The digital map database is the key component for the vehicle navigation system. The information of the road map, the landmarks, such as the government institutions and the famous buildings, and the life information, such as the hotel, park area, restaurant, gas station, and so on, also should be digitalized and stored in the digital map database. The 133 Design and Implementation of Vehicle Navigation Systems Figure 1. System platform basic representations for the digital road map are nodes, segments, and shape points. A node can be represented a cross point or an endpoint of a road. A segment can be represented the piece of road between two nodes. A shape points is placed in the segment, and is used to present the shape of the road. The location information of each node and each shape point, such as the latitude and the longitude, are stored in the digital map database. The landmark points can be used to denote the landmarks and the life information. The landmark points also contain the location information, such as the latitude, the longitude, and the address. These nodes, shape points, and landmark points are very useful for the map-matching process. In order to present the precisely road map, and provide more useful information, a lot of the shape points and the landmark points will be created in the digital map database. The computation cost of vehicle location and navigation will increase when the amount of these points increased. The hierarchical layer is introduced to improve the computation complex of vehicle location and navigation. These nodes, shape points and landmark points can be classified into different layers 134 b. according to their importance. For example, the nodes and shape points placed on the highways are in the upper layer, the nodes and shape points placed on the country roads are in the lower layer, the landmark points represents the airports, train stations, and harbors, are in the upper layer, and the landmark points represents the park areas, restaurants, and gas stations, are in the lower layer. The vehicle navigation system can retrieve these nodes and points according the drivers requirement. For example, the route planning process can only retrieve the nodes on the highways and arterials, and the map-matching process can retrieve the landmark points in the lowest layer when the vehicles is in the city. Moreover, the drivers can show the information of different layer on the screen upon their requirement. Location modules: The positioning devices and the map-matching function are two vital components in the location modules. The major functions of the location modules are to get the coordinates of a vehicle on the surface of the Earth from the position devices, obtain and show the precisely position of a vehicle on the map. The most common used positioning devices in the positioning Design and Implementation of Vehicle Navigation Systems c. devices can be classified into two groups. The relative devices, such as transmission pickup, wheel sensor and gyroscope, can estimate the approximate position data, such as the directions and travel distances. The absolute devices, such as GPS and compass, can report the precisely position information. Both these devices can be used on the vehicle navigation system. The map-matching function receives the location information of the vehicle from the positioning module, and retrieves the neighbor landmarks from the digital map library. Then, according to the location information of these landmarks, the map-matching function will obtain the placement of the vehicle relative to landmarks, or roads, and show the result on the map. The dead reckoning is the primitive technique used to determine the vehicle location relative to a reference point. When the starting location and the previous displacements are known, the vehicle position will be more reliable and accurate. The vehicle and distance traverse are widely used to estimate the changes in the position of the vehicle relative to the origin. The drivers also can input a query address into map-matching function and see the precisely position on the map. Route planning process: The route planning process is used to plan a path from the starting point to destination. The starting point can be the current position of the vehicle, or entered by the drivers from the user interface. The drivers also can select the landmark from the digital library and set as the starting point or destination point, or input the destination information from the user interface. Then, the route planning module will retrieve the road map from the digital library and plan the path from the starting point to destination. The planned route can be listed on the screen, and used by the route guidance module. The route planning d. process can be classified into the single vehicle route planning and multi vehicle routing planning. The single vehicle route planning can be solved by the shortest path algorithms, such as Dijkstra’s algorithm and Bellman-Ford algorithm. The single vehicle route planning is to determine the optimal route under the time or fuel constraints. The all-pairs shortest path algorithms, such as Floyd-Warshall algorithm and Johnson’s algorithm, can be used to solve the multi vehicle routing planning problem. The multi vehicle routing planning is to determine the optimal routes for all vehicles through the minimal total tour length subject to the time or fuel constraints. The computation complex of the route planning process is the critical problem. The hierarchical concept is introduced to improve the computation complex of route planning. The route planning process can only retrieve the nodes on the highways and arterials, and the highways and arterials segments, and determine the optimal route. In addition, the divide and conquer method can also be introduced to improve the complexity of the route planning process. The road map can be divided into several blocks, and the route planning process finds the path in each block. Finally, the route planning process assembles these block paths into a complete route from the starting point to the destination. The drivers can browse each block paths on the user interface. Moreover, the route planning process should provide the re-planning function, when the emergency event or congestion event occurs. Route guidance process: The aim of the route guidance process is to guide the drives along the planned path from the current position of the vehicle to the destination. The route guidance process should retrieve all the road information involved in the planned route from the digital library, and list that information, such as the road names, travel 135 Design and Implementation of Vehicle Navigation Systems e. 136 distances, turns, and landmarks, on the output devices. The drives can read that information before the trip. The route guidance process also can present that information to the drivers in real time while en route. The route guidance process should retrieve the current position of the vehicle from the location modules, and presents the proper guidance messages, such as turn-by-turn instructions, to the output devices. The route guidance process continuously retrieves the current position and direction of the vehicle from location the modules, continuously compares the position and direction with the planned route, and shows the turn message or some travel messages on the user interface or provide a series of voice announcements to warn the drivers. When the route guidance process finds the drivers off the planned route, a series of voice announcements should be provide to alert the drivers driven back to the planned route. The re-planning function provided by the route planning process also can be invoked to determine a proper path from the current position and direction to planned route, and the route guidance process guide the drives back to the original route. Moreover, the replanning function also can determine a new route from the current position and direction to destination. When the emergency event or congestion event occurs, the route guidance process should alert the drivers to plan the new route. Human-machine interface: The humanmachine interface provides a comfortable and effective human use interface between the users and vehicle navigation system. The drivers control the system from the input devices, and read the information from the output devices. The input and output devices can be classified into the visual display based interfaces and voice based interfaces. The drivers browse the digital map from the LCD f. or touch screen, and enter the command in the touch panels or touch screen. The drivers also can send the voice commands into the system by the microphone, and hear the announce message from the speaker. A good design of the human-machine interface can improve the driving safety. The design principles for a good human-machine interface were proposed in (Ligs et al., 1995). Wireless communication interface: The wireless communication interface is an optional component for the vehicle navigation system. The vehicle navigation system can provide route planning and route guidance without the wireless communication interface. However, the wireless communication interface can improve the driving quality. The wireless communication interface receives the current traffic information, such as emergency information and congestion information, from the traffic control center. Once these messages sent, the vehicle navigation system may alert to the driver. Then, the system can re-plan a new route based on that real time traffic information. The vehicle navigation system also can report the traffic information to the traffic control center. The traffic control center can get more precisely traffic information, determine the proper traffic control policy, announce that information to all vehicles, and improve the traffic quality. Moreover, the traffic information also can be exchanged between the vehicles without the traffic control center. This kind of network is referred as Vehicular ad hoc network (VANET), which a kind of wireless ad-hoc network. The VANET includes some special characters, such as an open peer-topeer network architecture, lack a central instance, each node is willing to forward data for other nodes, self-configuration and self-maintenance capabilities. Design and Implementation of Vehicle Navigation Systems system communication advance The system platform and system communication of the vehicle navigation system is shown in the Figure 1. The digital map database is the kernel component, which provides the information of roads and landmarks to the map-matching process, route planning process, route guidance process, and human-machine interface. The digital map database can update the real time traffic data received from the wireless communication interface. The map-matching process retrieves the roads and the landmarks from the digital map database. According that information and the vehicle information provided by the position devices, the map-matching process provides the precisely location on the map to the route guidance process and shows on the human-machine interface. The route planning process retrieves the roads, landmarks and real time traffic information from the digital map database, and receives the real time traffic information from the wireless communication interface, plans a path from the starting point to destination, and passes the path to the humanmachine interface and route guidance process. The route guidance process retrieves current position from the location modules, compares the position with the road information and real time traffic information in the planned route retrieved from the digital map database, and announces proper turn-by-turn message to the drivers driven along the planned path. The drivers also can browse the map from the digital map database, and use the route planning process to re-plan the route. The drivers can use the human-machine interface to report their known information to the traffic control center. ADVANCE (Advanced Driver and Vehicle Advisory Navigation ConcEpt) (Ligs et al., 1995; Boyce et al., 1994; Boyce 2002) is a project promoted by the industrial, government and academic of Illinois, USA in July 1991. The project test was completed in the end of 1996. The ADVANCE architecture is shown in Figure 2 (Zhao, 1997). The Traffic related function is placed on the traffic information center, and can provide history traffic information and static traffic profiles. The dynamic traffic information reported from the mobile navigation assistant (MNA) are aggregated by the traffic related function. The traffic information center (TIC) has many computers and devices to monitor the road network traffic, and report the traffic information to the MNA through the communication network. Both the TIC and MNA have the digital map database used to show the road map. The MNA is carried by the vehicle, and its architecture is shown in Figure 3 (Zhao, 1997). The navigation computer is the key component of MNA. The navigation computer gets the location information from the position devices, and retrieves the map information from the digital map database, which is a CD, and is read from the CD-ROM. The map-matching function, route planning function and route guidance function are provided in the navigation computer. The navigation computer shows the road map on the touch screen, and announces the voice message from the speaker. The navigation computer also reports the traffic data which receives from the position devices to the TIC, and receives the real time traffic information from the TIC through RF modem. Due to the computation complexity, the route planning function in ADVANCE only can consider the highway. The tourists are suitable for the planned route obtained by the ADVANCE, but the familiar drivers are not suitable. case sTudy Some vehicle navigation systems had been proposed in the past few years. Some of these implement architecture are illustrated in the section. 137 Design and Implementation of Vehicle Navigation Systems Figure 2. ADVANCE architecture Figure 3. Mobile navigation assistant architecture papago PaPaGo is mobile navigation software implemented by the Maction Technologies, Inc. in Taiwan (PaPaGo SDK). The platform of PaPaGo is WinCE, and the implement environment is Visual C++. PaPaGo can be installed on the Pocket PC, Smartphone, and PDA. PaPaGo can be divided 138 into four components. The map display component is the digital map database, the information retrieval component used to search the object from the map component, and the route planning component is to plan the path from the starting point to destination. The route guidance function and human-machine interface are integrated in the map display component. The position component Design and Implementation of Vehicle Navigation Systems of PaPaGo is GPS. Thus, the signal quality of GPS is very important for PaPaGo. Figure 4. VETRAC architecture VeTrac VETRAC (Thangavelul el al., 2007), as shown in Figure 4 (Thangavelul el al., 2007), is a vehicle tracking and navigation project based on the VANET. VETRAC does use the WiFi but not GPS to track the vehicles. VETRAC deploy some WiFi access point in the field, which can be used to receive the data sent from the mobile hosts, and also can be used to monitor the traffic flow and track the vehicles. The mobile computer with the client control panel and display map system is installed on the vehicle and uses the Mobile IP. The vehicle will report the traffic information to the vehicle tracking system and send some requirements to the navigation server through AP or VANET. As shown in Figure 5 (Thangavelul el al., 2007), the navigation server is the core component of VETRAC, and processes the traffic information sent from the vehicle tracking system, and response the user requirements. The navigation server also reports the location information to each vehicle, and stores the traffic information, digital road map, and vehicle locations into its database. Four servers are implemented in the navigation server. The connection server is used to manage the connections and the registries between the navigation server and the connection manager of the client control panels. The aim of the location server is to response the location query sent from the location manager of the client control panels. The traffic service can announce the traffic information retrieved from the client location database to the traffic information manager of the client control panels. The positioning server of the navigation server receives the traffic information sent from the vehicle tracking system, and updates that information stored in the client location database. The vehicle tracking system is used to aggregate the traffic information sent from the AP. When new vehicle enters, the vehicle tracking system will automatically report the positioning server of the navigation server. The drivers see the location information from the display map system of client control panel. dmrg Dual Mode Route Guidance System (DMRG) (Muffat el al., 1991; Kuhne el al. 1995) is a vehicle navigation system promoted by the car industries in the Europe. The system architecture of DMRG is shown in Figure 6. DMRG integrates the individual route guidance systems, and collective route guidance systems. The individual route guidance systems only plan a single vehicle route from the starting point to destination. The collective route guidance systems placed on the traffic control center, can collect all the traffic information, and provide the better planned router for each vehicle. The individual route guidance systems use the on-board computer to plan the path according to individual criteria (shortest 139 Design and Implementation of Vehicle Navigation Systems Figure 5. VETRAC software architecture Figure 6. DMRG system architecture way, quickest, simplest, etc). Thus, the several drivers use the same routing algorithm with the same criteria and destination will have the same planned path to follow. Some roads may become 140 congestion due to the case. In contrast, the traffic control center has each vehicle’s location and destination, the collective route guidance systems can plan different path for the vehicle with the Design and Implementation of Vehicle Navigation Systems same destination, and the traffic congestion can be reduced. The system architecture of DMRG is very similar to ADVANCE. The major difference is the route guidance function. ADVANCE only considers the individual route guidance systems, but the DMRG integrates the individual route guidance systems, and collective route guidance systems. conclusion In this chapter, we have shown the system architecture and system communication of the vehicle navigation system. The user requirement is described in the first. After system analysis, the modules of the vehicle navigation system are proposed, and the functions of each module are briefly illustrated. The general operations of the system communication are also described in the following section. At last, we elaborate some architectures of the proposed navigation system, and describe the difference of these systems. In recent years, human factors have been included into navigation systems in order to develop the intelligent transportation systems (Woodrow et al., 1998; Lee et al., 2005), and to improve the safety of drivers (Baldwin, 2002; Transportation Research Board, 2004). Summarizations these system architectures, the navigation systems still have a large scope for development. reFerences Baldwin, C. L. (2002). Designing in-vehicle technologies for older drivers: application of sensory-cognitive interaction theory. Theoretical Issues in Ergonomics Science, 3(4). doi:10.1080/1463922021000009029 Boyce, D. (2002). A Memoir of the ADVANCE Project. Journal of Intelligent Transportation Systems, 7(2). Boyce, D. E., Kirson, A. M., & Schofer, J. L. (1994). ADVANCE-The Illinois Dynamic Navigation and Route Guidance Demonstration Program. Advance Technology for Road Transport, IVHS and ATT. Frost & Sullivan, (2005). Strategic Analysis of the Taiwan Telematics and Infotainment Market. Green, P., Levison, W., Paelke, G., & Serafin, C. (1995). Preliminary Human Factors Design Guidenlines for Driver Information Systems. Technical Report No. UMTRI-93-21, Transportation Research Institute, The University of Michigan, Ann Arbor, MI. Kuhne, R. D., & Langbein-Euchner, K. (1995). Calculation of travel time savings by dual mode route guidance for the South corridor in the Stuttgart test field. Vehicle Navigation and Information Systems Conference. Lee, J., Forlizzi, J., & Hudson, S. E. (2005). Studying the effectiveness of MOVE: a contextually optimized in-vehicle navigation system. In Conference on Human Factors in Computing Systems. Ligs, J. F., & Bowcott, S. (1995). ADVANCE: Initial Deployment. Annual Meeting of ITS America. Muffat, M., Green, P., & Crosnier, S. (1991). European cooperation on dual mode route guidance—Perspectives for advanced research partners. Vehicle Navigation and Information Systems Conference. PaPaGO SDK. (n.d.). Retrieved from http://www.papagosdk.com/ Thangavelul, A., & Bhuvaneswari, K. Kumar’, K., SenthilKumarl, K., & Sivanandam, S.N., (2007). Location Identification and Vehicle Tracking using VANET (VETRAC). In IEEE International Conference on Signal Processing, Communications and Networking 2007. 141 Design and Implementation of Vehicle Navigation Systems Transportation Research Board. (2004). Transportation In An Aging Society: A Decade of Experience. Transportation Research Board. Woodrow, B., & Thomas, A. (1998). Dingus Human Factors in Intelligent Transportation Systems. Mahwah, NJ: Lawrence Erlbaum Associates. Zhao, Y. (1997). Vehicle Location and Navigation Systems. Boston: Artech House. 142 Section 4 Integrated Vehicular Application 144 Chapter 10 Vehicular Metropolitan Area Network Systems Architecture: The WiMAX Network Reference Model Cheng Hsuan Cho National Chung Cheng University, Taiwan, R.O.C. Jen-Yi Pan National Chung Cheng University, Taiwan, R.O.C. absTracT The WiMAX NWG develops a network reference model to serve as an architecture framework for WiMAX deployments and to ensure interoperability among various WiMAX equipment and operators. The network reference model envisions unified network architecture for supporting fixed, nomadic, and mobile deployments and is based on an IP service model. The authors introduce WiMAX network architecture, WiMAX network entry, mobility management, QoS functional elements, core network planning and accounting architecture in this section. However, all of them are significant in deploying WiMAX core network. The operator tries to reach the goals including system performance, reliability, and so on. On the other hand, the WiMAX operator should consider and balance such many variables in order to achieve a better situation. inTroducTion Based on WiMAX Forum Network Architecture definition (Stage 2: Architecture Tenets, Reference Model and Reference Points (Juo et al., 2008)) (Stage 3: Detailed Protocols and Procedures (WiChorus, 2009)), the WiMAX network reference model includes the logical functional entities and reference points. Figure 1 shows the WiMAX network reference model containing reference points, functional entities, and some important basic elements like Network Service Provider (NSP) and Network Access Provider (NAP). The functional entities contains Subscriber Station (SS)/Mobile station (MS), ASN(Access Service Network) and CSN(Connectivity Service Network). As in Fig 10.1, the network reference model depicts telecommunication’s operations. Access Service Network is managed by NAP (Network Access Provider). CSN is managed by NSP (Network DOI: 10.4018/978-1-60566-840-6.ch010 Copyright © 2010, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global is prohibited. Vehicular Metropolitan Area Network Systems Architecture Figure 1. WiMAX network reference model Service Provider). CSN provides IP connectivity, links out with the internet and combines with its Home-ASP (Application Service Provider). Basically, the overall architecture can be divided into three parts. The first part (left position in Figure 10.1) is MS/SS equipments, and it’s also called the end users. The second part (center position in Figure 10.1) is ASN. ASN supports wireless communication and air-interface to MS/SS. The third part is CSN where inner management servers locate (such as Home Agent, Location Register, AAA server and so on). Based on Figure 1, we give an introduction for these functional entities and reference points: ASN(Access Service Network): It links with WiMAX Customer Premises Equipment (CPE) e.g. MS, SS. Access Service Network (ASN) functions support WiMAX Layer2 connectivity, AAA messages forwarding, WiMAX NSP selection, Layer-3 link tunneling (BS-SS) and Radio Resource Management (RRM). The operations can be seen in the ASN-GW function. ASN-GW(ASN Gateway): As in Figure 2, we can observe the ASN-GW functions which contain reference points and functional entities in ASN. ASN-GW not only manages IP Data Forwarding in ASN, but also links with the other function entities inside or outside the ASN-GW. It tunnels data and packets to the suitable BS. ASN-GW Control plane handles all of the radioindependent control and includes authorization, authentication, and accounting (AAA), context management, profile management, service flow authorization, paging, radio resource management, and handover. Data plane feature set includes mapping radio bearer to the IP network, packet inspection, tunneling, admission control, policing, QoS and data forwarding. ASP(Application Service Provider): ASP provides basic network service and manages applications with IP-network. An application service provider (ASP) is a business that provides computer-based services to customers over a network. Software offered using an ASP model is also sometimes called On-demand software. CSN(Connectivity Service Network): The CSN contains several functional elements and tasks for supporting IP-Network connectivity to 145 Vehicular Metropolitan Area Network Systems Architecture Figure 2. ASN-GW reference model CPE. The following lists the major tasks in CSN network: 1. 2. 3. 4. 5. 6. 7. MS/SS IP address assignment. Internet access. AAA Server management. Tunnel supported (between CSN and SS). Accounting /pricing service. Users roaming across ASNs/ CSNs. Mobility management between MS and ASN. NAP (Network Access Provider): ASP provides basic network service and manages applications with IP network. Moreover, it provides radio link resource to one or more NSPs. One NAP can also contain one or more ASN network. For 146 example as Figure 10.1 shown, the NAP function contains two ASN networks. HA (Home Agent): HA supports mobility management for MIP registration, de-registration, and packet tunneling in CSN, and it helps CSN maintain the IP address table or location area, e.g. it records MS location where MS attached the serving BS. So CSN can use this information to tunnel packets to MS/SS from corresponding node (CN). In the network reference model, HA is located in CSN. FA (Foreign Agent): Foreign agent is a router serving as a mobility router for a mobile node (The definition is also in IETF-RFC-2002). A foreign agent works together with home agent to support Internet traffic forwarding. For the MS/ SS, it connects to the Internet from any location other than its home network. The home agent Vehicular Metropolitan Area Network Systems Architecture Table 1. Reference points R1~R8 Links Operation and Task R1 MS -- ASN Implementation as the same as IEEE 802.16e-2005 definition R2 MS -- CSN Authentication, authorization for MS, IP host configuration and mobility management R3 ASN -- CSN Supporting AAA, policy enforcement, and mobility - management capabilities (including tunneling and forwarding) R4 ASN-GW -- ASN-GW R5 CSN -- CSN R6 BS -- ASN-GW R7 In ASN-GW R8 BS – BS Used for MS mobility across ASNs For internetworking between home and visited network Implementation of intra-ASN tunnels and used for control plane signaling Coordination between data and control plane in ASN-GW Handover control, ensuring fast and seamless handover tunnels packets intended for the mobile node to a care-of address, which is either the IP address for the foreign agent, or an IP address acquired through some external equipment or server, such as Dynamic Host Configuration Protocol (DHCP) server. However, the foreign agent de-capsulates packets and delivers them to the mobile node. (Detailed MIP operations are described in RFC3344-IP Mobility Support for IPv4 (Mustafa Ergen, 2009)). LR (Location Register): Location Registers are database functions storing information typically used in the routing of signaling information. Typical examples include the HLR (Home Location Register) and VLR (Visitor Location Register) used in telecommunication. BS (Base Station): BS serves Subscriber Stations (point-to-multipoint), and provides SS with first-mile (or last mile) access to public networks. Besides, BS is directly connected to backbone networks (e.g Ethernet). The complement definitions of PHY/MAC layer are introduced in IEEE 802.16. It can also support radio resource management: call admission control, packet scheduling, power control and handover control. Base station has the local control of the network and gets assistance from ASN-GW for some features and implements the decision of the gateway for others. RP (Reference Point): In NRM, Reference point 1~8 are reference points between different functional entities. We conclude these and show their work as in Table 1. wimaX neTwork enTry The original network entry procedure is defined in WiMAX Forum stage2/3. The definitions are also described in 802.16e. However, the operation can be separated by the following operations as shown in Figure 3. nd&s (network discovery and selection) 1. 2. NAP discovery: NAP discovery is the first step of network entry. At first, an SS/MS detects available NAP(s) by scanning and decoding DL-MAP of ASN on detected channel to find NAP(s). SS/MS will range with the NAP if the NAP exists, then decode DL-MAP, UL-MAP, DCD and UCD. (The detailed operation is defined in IEEE 802.16 d/e) NSP discovery: The NAP may support more than one NSP. In Figure 4, if NSP Identifier Flag LSB in Base station ID equals to 1, it means there are more than one NSP in NAP. And also, MS acquires with NSP ID list. After this process, MS must discover each 147 Vehicular Metropolitan Area Network Systems Architecture Figure 3. WiMAX network entry MS indicates its NSP selection by attaching to an ASN associated with the selected NSP, and by providing its identity and home NSP domain in form of NAI (network access identifier). authentication 3. 4. NSP until all of these NSP are founded. On the contrary, if NSP identifier Flag LSB is equal to 0, MS will knows there is only one NSP service behind the NAP. NSP enumeration and selection: After NSP discovery, MS produces a list of available NSPs as discovered through NSP discovery in the available NAPs, MS uses the list select NSP by manual or automatic selection. ASN attachment based on NSP selection: Following a decision to select an NSP, an Figure 4. Base station ID format (Juo et al., 2008) 148 This protocol implements 802.16e security with IETF EAP framework. The protocol is used for service flow authorization, mobility management and policy control. AAA framework is based on pull model in which supplicant sends a request to ASN, and then ASN forwards it to AAA server. The AAA returns with appropriate response to ASN which sets up the service and informs the MS. The elements are depicted in Figure 5. The operation applies PKMv1 and PKMv2 by Extensible Authentication Protocol (EAP) to authenticate with user. PKMv1 is authentication for device. In contrast with PKMv1, PKMv2 is authentication for user and Device between CSN and MS. In the authentication flow, user Authentication is optional, while device Authentication is Mandatory. When the authentication procedures just only include Device Authentication or User Authentication, Single EAP can support the procedure. But when user and device both need Vehicular Metropolitan Area Network Systems Architecture Figure 5. WiMAX user authentication protocol authenticated, double EAP is needed, and each uses different AAA server (e.g. AAA servers is located in different CSN). configuration setup The procedure of Configuration Setup is after Authentication procedure. In order to support different network architecture (e.g. IPv4 or IPv6, network with or without DHCP), many configuration processes are proposed: ProxyMIPv4 (PMIP4), ClientMIPv4 (CMIP4), Proxy PMIPv6 (PMIP6) and Client MIPv6 (CMIP6). The “Client” means MS/SS support Mobile IP (MIPv4 / MIPv6) function. It doesn’t need other equipments forward or tunnel packet. Proxy Mobile IP (MIPv4 / MIPv6) means Proxy server is located in ASN which helps MS/SS in charge with registration and signaling MIP packets to HA. Therefore, although MS/ SS doesn’t have capability to supporting MIP, it can still roams in the wireless network by using Proxy-MIP server. data Transfer At this stage, MS/SS is able to use WiMAX core network. The user can use the data link to browse web-sites and transfer data. But when the user moving to another BS’s coverage, not only the attached WiMAX ASN-gateway will change, but also the IP address will change. We will discuss the mobility issue in the later paragraph about moving between ASN/CSN scenarios. wimaX mobiliTy managemenT The WiMAX network supports two types of mobility: (1) ASN-anchored mobility and (2) CSN-anchored mobility. ASN-anchored mobility is also referred to as intra-ASN mobility, or micromobility. In this case, the MS moves between two data paths while maintaining the same anchor foreign agent at the northbound edge of the ASN network. The handover in this case happens across the R8 and/or R6 reference points. ASNanchored handover typically involves migration of R6, with R8 used for transferring undelivered packets after handover. It is also possible to keep 149 Vehicular Metropolitan Area Network Systems Architecture the layer three connection to the same BS (anchor BS) through the handover and have data traversal from the anchor BS to the serving BS throughout the session. On the other hand, CSN-anchored mobility is also referred to as inter-ASN mobility, or macromobility. In this case, the MS changes to a new anchor FA, this is called FA migration. The new FA and CSN exchange signaling messages to establish data-forwarding paths. The handover in this case happens across the R3 reference point, with tunneling over R4 to transfer undelivered packets. Figure 6 illustrates the various possible handover scenarios supported in WiMAX. Based on different mobility scenarios (e.g. ASN-anchored mobility, CSN-anchored mobility), we can see Intra-ASN mobility and inter-ASN mobility between the WiMAX core network. WiMAX End-to-End Network Systems Architecture Stage 3 shows ASN Anchored Figure 6. Mobilityscenario with multiple ASN-GW 150 Mobility and CSN Anchored Mobility function and operation: asn anchored mobility ASN anchored mobility takes place when MS moves to a neighboring BS that connects to a different ASN-GW, which may be within the serving ASN or not. If the target BS belongs to a different ASN, the target ASN-GW will establish a R4 data path to the anchor ASN so that the MS can avoid data loss and the QoS could also be guaranteed. During the ASN anchored handoff procedure, the MS does not change its CoA. ASN Anchored Mobility divides ASN function into preparation phase and Action phase. 802.16e working groups defined MOB MSHO-REQ (when the network initiated, we called MOB BSHOREQ) message format for preparation phase and defined MOB HO-IND message format for action phase, some of preparation’s steps is optional. It Vehicular Metropolitan Area Network Systems Architecture can keep these steps on preparation phase until action phase. On the contrary, if we did these steps in preparation phase, the action phase leaves out the operation time that can reduce a part of handover latency. For ASN anchored mobility, traffic to MS comes from the anchor ASN. Figure 10.7 depicts the data flow of ASN anchored mobility. ASN1 is the serving and anchor ASN to MS. After handover, the packets from are CSN delivered from ASN-1 to ASN-2 via a R4 data path. The anchor ASN to MS remains the ASN-1. csn anchored mobility On the contrary, CSN anchored mobility involves in anchor ASN relocation, which means anchor ASN is changed after CSN anchored mobility. Figure 8 describes the data flow after CSN anchored mobility. The packets directly send from CSN via the new R3 link to ASN-2, unlike in ASN anchored mobility the traffic comes from ASN-1. ASN-2 is the new anchor ASN for MS. Based on these mobility scenarios, we make a simple conclusion and comparisons in Table 2. ho Function network Transaction The handover (HO) allows MSs to handover between neighboring BSs while moving across the corresponding coverage areas. Furthermore, the mechanism can be used by BSs to trigger a HO in order to optimally balance the traffic load of cells within a network. The basic HO function transaction is shown in Figure 9. HO function contains serving HO, relaying HO and target HO. 1. At first, the serving HO function initiates an HO network transaction by sending HO_Req. There can be only one serving HO function for any given HO network transaction. After Figure 7. ASN anchored mobility scenario 151 Vehicular Metropolitan Area Network Systems Architecture Figure 8. CSN anchored mobility scenario Table 2. Comparison with ASN- anchored mobility and CSN- anchored mobility Mobility type Handoff latency ASN- Anchored Mobility CSN- Anchored Mobility Faster – handoff process without through FA Slow– handoff process through FA FA change or not ASN-GW’s FA does not change, and the original FA tunnels packets to new BS. CSN changes FA. MS re-registers with CSN and it’s HA with MIP registration procedure. Then, new FA will forward packets. From HA viewpoint When MS move to another ASN, HA doesn’t knows that. MS will change its CoA on HA’s mapping table when it re-registers. Backhaul Link Efficiency -- Better 2. 152 receiving HO IND from MS, serving HO function the serving HO function confirms HO to only one target HO function by sending HO Confirm messages. The target HO Function responds to the HO network transaction with HO_Rsp. There can be one or more target HO functions for an HO network transaction. 3. Serving HO Function and target HO functions communicate either directly or with assistance of one or more relaying HO function. If the serving and target HO functions cannot communicate directly for any reason, the relaying HO function takes care of delivering the relevant information to the corresponding target HO functions. A single HO primitive (e.g. HO_req) that is sent Vehicular Metropolitan Area Network Systems Architecture Figure 9. HO function transaction from the serving HO functions may contain information relevant for several target HO functions. In this case several behavioral policies might be applied. Qos FuncTional elemenTs Under the IEEE 802.16 specification, a SS could be associated with a number of service flows characterized by QoS parameters, and QoS framework for the air interface. The specification also defines combined scheduling scheme, Resource allocation and admission control. Besides, the WiMAX forum also describes QoS framework for WiMAX Network service, as shown in Figure 10. The WiMAX Forum has specified a framework for Service Management and QoS. Service Flow (SFA) and Service Flow Manager (SFM) are the entities that act as policy decision and enforcement points respectively for Service Management and QoS. The following figure depicts the policy framework for WiMAX networks: As service flow provide a particular QoS according to the QoS parameter set defined for that service flow. The following mechanisms of the standard are defined to ensure the end-to-end QoS: 1. 2. 3. 4. 5. Pre-provisioning of the service flow parameters. Signaling function to dynamically establishing the QoS enabled service and traffic parameters. Utilization of MAC scheduling and QoS traffic parameter for uplink service flow Utilization of QoS traffic parameter for downlink service flow. Grouping of service flow properties into name Service Classes. Based on Figure 10, we give an introduction for these functional entities: 1. SFA (service flow authorization): Is responsible for evaluating any service request against the subscriber’s QoS profile. SFA logical entities in the ASN. In case the user QoS profile is downloaded from the AAA into the SFA at network entry phase, the SFA is responsible for evaluating any service request against user QoS profile. The SFA 153 Vehicular Metropolitan Area Network Systems Architecture Figure 10. QoS functional elements (Juo et al., 2008) 2. 3. 154 also performs ASN-level policy enforcement using a local database and an associated local policy function (LPF). The LPF can also be used to enforce admission control based on available resources. SFM (service flow management, SFM): Is responsible for the creation, admission, activation, modification, and deletion of 802.16 service flows. It consists of an Admission Control (AC) function and the associated local resource information. PF (policy function, PF): Resides in both home and the visited network, comprising 4. 5. their respective databases. The databases include general policy rules as well as application-dependant policy rules. AF (application function, AF): Hosts the service logic and communicates the application level session information to PCRF such as classifiers identifying service flows, on which policy control and differentiated charging is required. AAA Server: Holds the subscriber’s QoS profile and the associated policy rules per subscriber. At first, when the user enter WiMAX network,AAAserver will download Vehicular Metropolitan Area Network Systems Architecture users QoS parameters for Authentication and Authorization. After that, AAA server can also release these QoS parameters to PF that PF can decide how to handle the message form AAA server. wimaX core neTwork planning We believe that planning and modeling are important things for build a good network structure. In this section, we suggest some factors needed to consider in planning a telecommunication network - 1. Network deployment 2. Quality of Service management 3. Mobility management 4. Network management. Based on these aspects, we’ll focus on those managements and give more details: 1. Network deployment: Network deployment is the first consideration for telecommunication/wireless communication operators. The WiMAX operator needs to figure out the capital expenditure and operation expenditure on their deployment scenario. A. CAPEX: The capital expenditure (CAPEX) includes the costs of wireless end device and WiMAX equipment (e.g. BS/SS/RS), backhaul link equipment, server and edge equipment. All of the costs when we set up our environment are contained in CAPEX. For example, spectrum license is included in CAPEX, many WiMAX equipments (802.16e BS/SS) are based on licensed bands (2.5G/3.5G), but some are based on un-licensed bands. In licensed bands, we need to consider how much bands that we need to cover and how much cost we need to pay for these bands which are differ from continuous or non-continuous bands. The first WiMAX-certified products will be operating in the licensed 3.5 GHz frequency band, followed by systems for both the 2.5 GHz licensed band as well as the 5.8 GHz license-exempt band. The important thing is to add all of this payment in our CAPEX. B. OPEX: An OPEX is an on-going cost for running a product. For larger systems like businesses, OPEX may also include the cost of workers and facility expenses such as rent and utilities. In the WiMAX core network structure, OPEX are costs associated with the operation and maintenance of an income producing property. It includes accounting expenses, advertising, office expenses, supplies, attorney fees, insurance and property management. C. Matching data density requirements to base station capacity (Akyildiz et al., 1999) D. For WiMAX deployment scenarios, data density is an good metric for matching base station capacity to market requirements. Demographic information, including population, households, and businesses per sq-km or per sq-mile, is readily available from a variety of sources for most metropolitan areas. With this information and the expected services to be offered along with the expected market penetration, data density requirements are easily calculated. This 6-step process is summarized in Figure 11. 2. QoS of network management: The 802.16 standard provides some QoS designs in order to achieve service quality demands, such as voice and video quality. Also, ISPs have their own QoS management schemes to provide different fees for service items, which means the provided solutions can be customized. Therefore, adequate collocation of service levels and charges must be considered so as to satisfy customers with fine revenue. 3. Mobility management: A user can move from the serving BS’s coverage to another target BS’s service coverage. In this scenario, user’s move involves handover processes. In fact, Handoff processes consist of three phases: handoff decision, radio link transfer and channel assignment (WiMAX Forum, 2007). Considering in handoff decision, several goals are important while designing and evaluating mobility management procedures: A. Small signal overhead over the air interface 155 Vehicular Metropolitan Area Network Systems Architecture Figure 11. Determining market driven capacity requirements (Akyildiz et al., 1999) Table 3. Factors of handoff decision Factors Network provider User B. C. D. E. Details of factor Bandwidth Available bandwidth of a network Coverage Coverage of a network Related service Academic/ enterprise/ commercial Connections Support numbers of user Mobility Support mobility management or not Network structure WLAN/3G/3.5G/WiMAX Application The application running on mobile node(MN) Power Power consumption of MN Price Offers service’s fee to ISP or network provider RSS Received Signal Strength from base stations Bandwidth The minimum bandwidth that user needed Packet loss Packet loss of the connecting network Low handover latency Minimum packet losses during handover Least handover failure Efficient use of network resources For the Network provider and user provider, those are many factors that we can consider in. However, we conclude from these observations. Table 3 shows the factors on handoff decision (WiMAX Forum, 2007): 4. Network management: While the network is setup, the operation and management of WiMAX network are required to handle by WiMAX operator. Those are the key points to maintain network service. We also divide it into the following parts in Table 4. 156 wimaX accounTing archiTecTure How to account and price in Wimax network? 802.16e working groups defined AAA- Authentication, Authorization and Accounting, and those are a standard for security pricing system. About the Authentication, we are already discussed before. It implements security procedure with IETF EAP framework. After that, WiMAX operator accounts for MS/SS based on some account procedure. Accounting architecture is shown in Figure 10.12. The figure shows network elements for both postpaid and prepaid services. The figure also shows the network elements for Hot-lining support and negative volume count for ASN. A Vehicular Metropolitan Area Network Systems Architecture Table 4. Factors of network management Content Entity management Content analysis BS/ SS/ Router/ Switch management, and so on. Integration Integration WiMAX network and backbone network or others network service (e.q PSTN, 3G) Inventory management Includes network resource as IP assignment and account for usage of bandwidth Network Planning Network service Coverage Ratio, radio resource and related network management Charge user’s account and deal with permission of user Billing service Manage and price service level agreements (SLAs) for differentiated services that use a simple upper bound for the effective bandwidth of the conforming traffic as a proxy for resource usage. customer relationship management (CRM) Good running of customer relationship management give an insight for firms to understand the market trend so that firms can adapt the changing environment and anticipate the future needs (e.g. web-site, customer service, maintenance) Data mining It can be used to uncover hidden patterns in data that have been collected (e.g. user’s preference) description of each entity is provided in the following sections. NWG defined the Accounting Architecture (Juo et al., 2008). As shown in Figure 10.12, it contains NAP, Visited NSP and home NSP. All of the operation and data flows are based on RADUS. The architecture defined offline (post-paid) accounting, QoS-Based accounting, and online (prepaid) accounting. These include the delivery of information for the purpose of billing (both prepaid and post paid billing) and information that can be used to charge activity by both the home NSP and visited NSP. These accounting procedure are introduced as following: Figure 12. WiMAX accounting architecture 157 Vehicular Metropolitan Area Network Systems Architecture offline (post-paid) accounting online (prepaid) accounting We describe the off-line (post-paid) accounting procedures and the Usage Data Records (UDRs) in this part. The Serving ASN must merge radio specific parameters (number of bytes/packets transmit at the BS) called Airlink Records with IP network into the Usage Data Records (UDR). After that, the Serving ASN uses RADIUS accounting messages to send UDR information to the home RADIUS Server (via the visited AAA server if the subscriber is roaming). The Visited and Home RADIUS server should be support the Accounting-Request records as specified in (WiChorus, 2009) specification. Based on the UDRs, Home AAA server can use the records manage and account for MS. This section describes the online (prepaid) billing procedures in the WiMAX core network. The prepaid packet data service allows a user to purchase packet data service in advance based on volume or duration Account status is stored on a prepaid server (PPS) that is located in the user’s home network and accessed via the HAAA server. To provide service to roaming prepaid users, the visited ASN or CSN needs to support the prepaid service and the local and broker AAA servers need to forward the new prepaid accounting attributes transparently to and from the home AAA server. The HAAA server and the prepaid server could be collocated or could be separate entities. However, the architecture is a definition for accounting procedure. Mostly, the operator combines the online, offline, and QoS based accounting method in their service, and they use the UDRs to charges for subscribers. For example, the WiMAX operator needs to price for their subscribers. For simply, offline and QoS accounting are pricing schemes for web surfing or downloading/ uploading of each subscriber. We denote it by application service fee. Conversely, most operators charge the basic service fee in every month, and come together with the service fee that we called basic service fee. However, by using two -part tariff strategy to subscriber is not always better. Depending on different strategy of WiMAX operator, the pricing scheme can be changed or modified. Regarding the accounting methods, the important thing is that WiMAX operator should forecast the profit of a broadband access system in MRT environment. Qos-based accounting The QoS-based accounting is based on user’s SLA. For the basic service, WiMAX BS supports the 802.16 defined QoS classifications. But when user bill payment for ensuring connection/bandwidth thresholds, the UDRs are also includes these parameters for discount if there is a mistake for wireless connection .for example, if the user chooses the high bandwidth scheme, the Home AAA server in CSN should support this pricing scenario for bandwidth accounting. On the other hand, based on different classifications, the QoS-based accounting supports different applications. For example, when we use the FTP, WiMAX operator may consider user’s FTP’s UDR and charge activity with the amount of FTP session’s packet. The scheme may also consider the time of user’s connection. For example, WiMAX operator focused on VoIP application and Video on demand (VOD) application and charge with the time of these applications. Based on these requirements, the Visited RADUIS, Home RADIUS, ASN-gw and BS should support the application’s accounting-Request records. In this way, QoS-based accounting can be used in WiMAX core network. 158 reFerences Akyildiz, I. F., Ho, J., & Wang, W. (1999). Mobility Management in Next-Generation Wireless Systems . Proceedings of the IEEE, 87(8), 1347–1384. doi:10.1109/5.775420 Vehicular Metropolitan Area Network Systems Architecture IETF Network Working Group. (2008). RFC3344IP Mobility Support for IPv4. Retrieved from http://www.faqs.org/ rfcs/rfc3344.html WiMAX Forum. (2007). WiMAX End-to-End Network Systems Architecture Stage 3: Detailed Protocols and Procedures, Release 1.1.0. Juo, C. S., & Pan, J. Y. (2008). Software Agent Framework for Dynamic Handoff Decision. In APCC 2008, Akihabara, Tokyo, Japan. Ergen, M. (2009). The Access Service Network in WiMAX: The Role of ASN-GW. WiMAX Forum. (2007). WiMAX End-to-End Network Systems Architecture, (Stage 2: Architecture Tenets, Reference Model and Reference Points), Release 1.1.0. WiChorus, Inc. (2009). Retrieved from http:// www.mustafaergen.com/asn_gateway.pdf WiMAX Forum. (n.d.). Wimax deployment consider for fixed wireless Access in 2.5GHz and 3.5GHz licensed bands. 159 160 Chapter 11 Interworking of IP Multimedia Subsystem and Vehicular Communication Gateway Wei-Kuo Chiang National Chung Cheng University, Chiaya, Taiwan, R.O.C. An-Nie Ren National Chung Cheng University, Chiaya, Taiwan, R.O.C. Abstract In recent years, more and more people dream of experiencing various IP-based multimedia application services when they are driving through their car. However, those multimedia devices in the car may use different communication protocols such as X.10, Havi, Jini, UPnP and SIP. In order to provide a variety of IP-based multimedia services to those users in the car, the authors mainly investigate the issue of interworking between IP Multimedia Subsystem (IMS) and telematics of the vehicular industry. A service-integrated platform, Open Service Gateway Initiative Service Platform (OSGi SP), has been proposed to act as a Residential Gateway (RGW) and to administer the communication between the vehicular environment and Internet. Besides, a Home IMS Gateway (HIGA), which can be implemented on a NGN RGW, has been developed by Home Gateway Initiative (HGI) since 2005 to collect the relevant information of in-car users, devices and services and to manage the IMS sessions for the in-car devices that do not support IMS functions. With these techniques, the users can enjoy their digital life by interacting with the home/vehicular network from anywhere. Introduction Vehicular network has emerged as a hot research issue recently. More and more people want to experience a variety of IP-based multimedia application services when they are driving through their cars. However, those devices on the car may use DOI: 10.4018/978-1-60566-840-6.ch011 different communication protocols. Due to the lack of interoperability between all devices on the car, a service-integrated platform is needed to provide the characteristic of interoperability. On the respect of home network environment, it will encounter this problem as well. With the wide spreading of the digital home products, Home Networking has gradually been becoming a popular issue in recent years. The Home Networking is a way to manage all Copyright © 2010, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global is prohibited. Interworking of IP Multimedia Subsystem and Vehicular Communication Gateway kinds of network devices or home electrical appliances; it allows the users to operate and manage theses networking electrical appliances through a unified interface or platform. These electrical appliances represent all kinds of facilities that have specified functions and could be controlled or administered by using specified network media or technology, e.g., X.10 device, HAVi device, Jini device and UPnP device. One of the goals of the Home Networking is to control the in-house electrical appliances from the external network by using hand-held devices (such as mobile phone, PDA) to connect to the home gateway through the IP network, and thus to control the electrical appliances inside the Home Networking (e.g., lamps, air-conditioner, TV). Besides, if there happens any status changing to the in-house devices, it (the networking system) could notify the user immediately. In addition to the Home Networking technology, we could also treat a car/vehicle or any public transportation as a mobile home network to integrate the home network and telematics of the vehicle industry. In order to fulfill these goals, there are yet some issues necessary to be discussed, e.g., security, mobility, the interoperability among the communication protocols, etc. In recent years, because the growth speed of the mobile devices and in-house network devices becomes very fast and the communication protocols and home network/vehicular communication technologies have undergone diversified changes, these have made the intercommunication among the-mentioned devices to be more difficult. For example, the communication protocols adopted by the in-house/in-car devices could probably be UPnP, IP and SIP; therefore, an Open Service Gateway Initiative (OSGi) has been proposed to construct a service-integrated platform among the Home Networking, vehicular communication environment and the Internet. Because the SIP communication protocol possesses the capabilities of security mechanism, event notification, media streaming and mobility management among the terminals, so, concerning the applications of remote accessing the home/ vehicular network, we will discuss how to support SIP on OSGi platform, how to utilize SIP as an unity way for message control, and we will further probe and discuss the mapping of control messages between SIP and UPnP in order to solve the interoperability problems among the devices that adopt different communication protocols. Since the IP-based network technology has become more popular, the future communication network will move toward the service integration of the all-IP network and use the IP Multimedia Subsystem (IMS) as a core network to provide a variety of multimedia services. Therefore, integrating the IMS and the vehicular communication environment will be a developing trend for future network service providers. A user could then monitor the in-car devices or access the data in the multimedia storage device through his mobile device in anytime and anywhere, this will bring the user a richer experience in vehicular communication network. relaTed works sip (session initiation protocol) SIP (Session Initiation Protocol) (Rosen et al., 2002) is a communication protocol developed by the MMUSIC (Multiparty Multimedia Session Control) task group of IETF (Internet Engineering Task Force). SIP is an application layer protocol for session control and signaling control, it could be used to initiate, modify and terminate sessions. The SIP application range is very extensive, including voice and video calls over Internet, video conferencing, presence service, event subscription/ notification and instant messaging. In November of year 2000, SIP has been accepted by (The Third Generation Partnership Project [3GPP], 2009) to become the protocol for conveying communication control _ignaling and has been applied in the IMS (IP Multimedia Subsystem) infrastructure of 161 Interworking of IP Multimedia Subsystem and Vehicular Communication Gateway the Next Generation Network (NGN). This SIP protocol could be used to establish and administer the sessions in real time among the users, which includes the application services such as text, voice, picture, video and interactive gaming, etc. It could also invite other users to join in an established communication session to provide a multicast conference service. SIP is similar to HTTP (Hyper Text Transmission Protocol), both are text-based communication protocols; SIP URI (SIP Uniform Resource Identifier) address is exactly like an e-mail address, e.g., Alice@example. com.In the following, we will simply introduce the fundamental elements of SIP: • SIP user agents: User Agents are the terminal equipments in the SIP network, which could be SIP phone or the SIP client software in a PC; it includes User Agent Client (UAC) and User Agent Server (UAS). UAC is responsible for generating requests, and UAS is responsible for generating the corresponding responses with respect to the received requests. Each Figure 1. Architecture of SIP message flow 162 • • • SIP User Agent will contain both UAC and UAS functions. SIP proxy: SIP Proxy is in charge of passing the received request to another SIP component or SIP UA. When a SIP UA sends a request, this message will not be directly sent to the SIP UA at the destination terminal, it could be passed through multiple SIP proxies before the request message will be sent to the terminating SIP UA. Registrar server: Registrar Server is responsible for processing the SIP UA registration, and it is used to manage the certain specified service as well as to update the location information for the SIP UA. Redirect server: When the Redirect Server receives a request from SIP UA or proxy, it will return with a Response of 3xx to inform the SIP UA or the proxy that this request should be directed to another SIP component. The SIP session establishment flow chart is shown in Figure 1. Firstly, the terminating UA Interworking of IP Multimedia Subsystem and Vehicular Communication Gateway must register with the Registrar Server; the registrar will then save the UA location information into the Location Server. After that, the originating UA will send the request to the outgoing proxy; the proxy will make an enquiry against the DNS server for the incoming proxy address of the terminating UA, and then pass this request to the incoming proxy. The incoming proxy of terminating UA will then interrogate the Location Server for the terminating UA address, and pass this request to the terminating UA. When a session has been established, the caller and the callee could communicate by using RTP (Real-time Transport Protocol) or other transmission protocols. ims (ip multimedia subsystem) The Next Generation Network (NGN) has integrated all types of heterogeneous networks to provide the diverse, novel and personalized multimedia application services. By observing the past telecom network development, we could see that, from 2G/2.5G to 3G/B3G and even the 3.5G/4G, the service technology development has been evolving gradually from the traditional voice communication service into the nowadays all-IP network framework that uses the IMS (IP Multimedia Subsystem) as its core; the system provides the multimedia services such as video phone, conference, VoD, OMA service enablers, etc. The IMS (3GPP TS 23.228, 2008; 3GPP TS 23.218, 2008) (see Figure 2) formulated by 3GPP adopts SIP protocol as the communication basis, it is a multimedia communication platform that provides functions of session management, security, mobility, QoS and charging. IMS is widely accepted as a network framework core that could realize the Fixed Mobile Convergence (FMC) and Triple Play Services and provide strong service supports. The Call Session Control Function (CSCF) of IMS is the core component of IMS. Its purpose is to process the signaling control for session set-up between the users or between the user and the server; it includes completing registration, basic call control, SIP signal route control, service trigger, etc. According to functional difference in IMS, CSCF could be classified into Proxy Call Session Control Function (P-CSCF), Interrogating Call Session Control Function (I-CSCF) and Serving Call Session Control Function (S-CSCF). The role of P-CSCF is just like a proxy server that is responsible for forwarding the SIP message in IMS. P-CSCF also acts as an User Agent (UA) in abnormal condition; I-CSCF is mainly in Figure 2. 3GPP IP multimedia subsystem 163 Interworking of IP Multimedia Subsystem and Vehicular Communication Gateway Figure 3. S-CSCF service triggering in IMS charge of interrogating Home Subscriber Server (HSS) and selecting a S-CSCF for the user, then forwarding the SIP message to the S-CSCF; the role S-CSCF plays is a registrar, it also controls the execution of the communication (call) and service triggering. Home Subscriber Server (HSS) is the main database of 3G telecom system. It is in charge of storing the criteria of service triggering for the user subscribed services. Application Server (AS) is responsible for executing all kinds of service logics, such as call forwarding, call waiting, voice mail, etc. In addition, AS also provides SIP-based service such as the services formulated by (Open Mobile Alliance, 2009) ; those include Multimedia Message Service (MMS), Push to talk over Cellular (PoC), mobile location service, etc. It is able to provide the user with more extensive multimedia services. And, S-CSCF could communicate with AS that supports SIP interface by going through IP Multimedia Service Control (ISC) that adopts SIP as the communication protocol. The IMS service triggering flow has been defined in 3GPP TS 23.218 (see Figure 3). The initial requests sent by the user will be cross-compared to initial Filter Criteria (iFC) that corresponds to the 164 added-value service subscribed by the user within the service call control component (S-CSCF). If they are conformable, then the application server (AS) should be added into SIP call control signaling path; AS will monitor the call session set-up status and decide whether to execute the addedvalue service subscribed by the user. osgi (open service gateway initiative) The full name of OSGi is (Open Service Gateway Initiative, 2009). The generally mentioned OSGi represents two meanings; it could represent the OSGi Alliance organization and could also be the service specification - OSGi service platform. OSGi service platform is mainly developed basing on Java language; OSGi is a standardization organization founded by Sun, IBM, Ericsson and others in March, 1999. Its main goal is to provide a complete solution for point to point service delivery between remote service provider and local devices. Therefore, OSGi Alliance has defined a open platform that allows the application program and its value-added service of the remote service provider to be downloaded at any time to the Gateway nearby the user according to the user Interworking of IP Multimedia Subsystem and Vehicular Communication Gateway Figure 4. OSGi framework layering demands; and it could be automatically installed. The gateway is a common device used to connect the home/vehicular network to Internet. Observing from the framework view, OSGi could be decomposed into three componenets: Framework, Bundle and Service. The framework is constructed on Java virtual machine and the application program executed on the framework is called bundle; the interface service provided by or required by each bundle is called a service. Bundle could be downloaded from the remote service provider and be automatically installed on the framework; after installing, the said bundle could then be executed. Bundle will register the service it provides in the OSGi platform and bundle could also make a request upon OSGi platform for the services provided by other bundles. Framework is an integral information service platform; its main function is to provide bundle with an environment of execution and dynamically to adjust the bundle life cycle. When a bundle has been installed, it will register the service it provides at the framework; however, when a bundle has been suspended, the framework will dynamically remove the services registered by the bundle. The structure of framework could be divided into several layering; they are respectively security layer, module layer, life cycle layer, service layer, and etc.; as shown in Figure 4. Bundle is an application program on an OSGi framework, it could be activated and executed by the OSGi framework. Since it is an application program, to define the specified file format is required. An OSGi bundle is a Java Archive (JAR) file; and the componenets included in each JAR file are respectively Java class, Activator class, Manifest header and some resource files (such as HTML web page or some graphic files); in which, the Manifest header mainly contains the description of that bundle information and will formulate some specifications such as Import-Packet, Export-Packet, Bundle-Activator, Import-Service, Export-Service, etc. Bundle life cycle are mainly managed by the OSGi framework, its life cycle could be divided into six states: INSTALLED, RESOLVED, STARTING, ACTIVE, STOPPING and UNINSTALLED. The flow chart of the bundle life cycle is shown in Figure 5. After an OSGi bundle has been installed on the framework, the services it provides could be presented by a service. A service must be a well-defined interface service. When each bundle provides a service, the framework will 165 Interworking of IP Multimedia Subsystem and Vehicular Communication Gateway Figure 5. Bundle life cycle reserve a corresponding service reference; and, when another bundle needs the service provided by this bundle, an interrogation could be made upon framework through LDAP (Lightweight Directory Access Protocol) to acquire that service. Inside the framework, an effective operation and application shall be intercommunicated through a series of services. Concerning the services presently defined by the OSGi specifications, those include: standard services and custom services, the two are shown in Figure 6. Standard services are provided by framework itself; comparing to standard services, custom services belong to the part that are defined and developed by the vendors on their own. Each vendor could develop its own bundle and the corresponding service to segment the market of the product. After the above-mentioned introduction, we could know that OSGi possesses the component administration function and allows the user to proceed with administration on components from remote site; each OSGi framework will even monitor the privilege of each bundle and differentiate the execution of each bundle. Based on the viewpoints mentioned above, we may realize that OSGi provides a safe environment for software execution. If each bundle has enough privilege, and the bundle could access the service provided by the counter-party bundle among one another, this will be helpful to the modularized program design. OSGi itself provides automatic bundle download, dynamic installation, registration and bundle updating; in which, it is not necessary to re-activate Java virtual machine and makes the operation easier and more user-friendly. For more detailed OSGi, one could refer to the specifications published by OSGi Alliance. Figure 6. OSGi standard services and custom services 166 Interworking of IP Multimedia Subsystem and Vehicular Communication Gateway Release 4 has been developed to be the current version, those who are interested in could visit OSGi Alliance official web site to download. upnp (universal plug and play) Universal Plug and Play (UPnP, 2009) is a protocol suite formulated by UPnP™ Forum founded in October, 1990. The protocol vision is to establish a seamless connection between the home network (data sharing, communication and entertainment) and the all kinds of devices within the enterprise network, and to simplify the practice of relevant network. In brief, it is expected that as long as each device (e.g., computer, TV, fridge, air-conditioner, alarm clock or lamps, etc.) has been connected to the network, all the already on-net devices could sense that there is a new device added in, these devices could then communicate one another and could be directly operated and controlled; no configuration is required for any device; one could fully enjoy the advantage of Plug and Play. For example, assuming one has bought a printer at home, if you expects this printer be shared and be used by all in-house computers, you will have to install this printer and set it to be sharable by all. After that, you will also have to install this network-shared printer on other computers; will it cause too much trouble? However, if we adopt UPnP as the intercommunication bridge for all in-house devices, as long as both the home printer and computer support UPnP, the user only needs to hook up the newly-bought printer, then, all in-house computers will notice that there is one printer available; no any configuration is required and the printer becomes usable. This is exactly the vision that UPnP wishes to realize. Making the configuration of the home network environment to be simplified could further enhance our living quality. UPnP is a technology realized on an open point-to-point IP network of Plug and Play. It is a type of web-based communication protocol; and, it adopts the present widely used standards like TCP/IP, UDP, HTTP, XML and SOAP to construct the UPnP platform. UPnP features also include: • • • • • Media and device independence: UPnP technology could be applied on many transmission media including Ethernet, IrDA, RF (Wi-Fi or Bluetooth), IEEE 1394 and Power lines. User interface control: UPnP technology has enabled the device vendors to control the device and to inter-communicate among one another through the web browser. Operating system and programming language independence: Any operating system and program language could be adopted to construct UPnP products. Program control: UPnP architecture also enables the conventional application program control. Extensibility: Each UPnP product can have device-specific services be layered on top of the basic architecture. The basic components that the UPnP consists of are Service, Device and Control Point. Service is the minimum control unit in UPnP; service provides operative actions and a set of status variables to record the current status of this service; device represents an UPnP device, the device is the facility that includes services. For example, a DVD player provides video-broadcasting service; the control point is able to control the device found in the UPnP network. In the following, we will give a simple introduction to UPnP protocol stack as shown in Figure 7: • • HTTPU/HTTPMU: These two communication protocols are extensions of HTTP. They use UDP/IP to transmit the data, and they are used by SSDP as well. SSDP: Its full name is Simple Service Discovery Protocol; it is a communication protocol built in HTTPU/HTTPMU; it 167 Interworking of IP Multimedia Subsystem and Vehicular Communication Gateway Figure 7. UPnP protocol stack defines mainly how to make the services be found out; it includes how the control point could find out what services exist on the network and access to the relevant information about these services; that the device itself claim what services it could provide is also completed through this SSDP protocol. • SOAP: Its full name is Simple Object Access Protocol. It mainly defines how to use XML and HTTP to execute the remote procedure call. • GENA: The full name is Generic Event Notification Architecture; it is mainly adopted to process the transmitting and receiving of subscription and notification messages. As to the communication between the device and control point, it could be separated as the following six phases: • 168 Addressing: It is the initialization step for a device to add into the UPnP network. In order to intercommunicate with the components in UPnP network, each control point and device must have an IP address. • • The way to get an IP address could be done by using DHCP or using auto IP to get the network address. Discovery: It allows the control point to find out the service of its interest. When a device is added into the network, that the device will broadcast its services to the control point in the network; when a control point adds into the network, it will search the device of its interest in the UPnP network. Using SSDP via HTTP/HTTPMU completes all these actions. Description: When a control point has found out the device from the step of Discovery, because the control point has pretty limited knowledge about that device, if it intends to further understand the device function or to interact with that device, it will use the device XML address obtained in the Discovery step to acquire the XML document of the device description. This document includes the device name, serial number and manufacturer, the actions and status variables provided by the device and the URL web address for controlling this device. Interworking of IP Multimedia Subsystem and Vehicular Communication Gateway • • • Control: When a control point intends to control a device, it will transmit a proper control message to the service URL according to the information acquired from Description phase, this control message is expressed in XML format by using SOAP protocol. When a device receives this message, it will act to change the corresponding variables and then forward back to the control point; if this action fails, it will return an error code. Event: A control point could subscribe the status variable of its interesting device. When the status on a device changes, the device will issue an Event message, and use GENA protocol to pass this message back to the control point. This Event message is also expressed in XML format. Presentation: If that device has a web GUI for web page presentation, then the control point will load that page into the browser, and then the control point could inspect the device status and control the device from that web page. The above is a simple introduction of UPnP; it describes the operating environment and features of UPnP, UPnP protocol stack and every single phase when being used. dlna (digital living network alliance) DLNA is the abbreviation of (Digital Living Network Alliance, 2009). It is an alliance organization founded in 2003, which consists of the vendors from consumer electronic products, mobile phones and computers. The organization goal is to establish a set of standard specifications that could allow the video/audio facilities of different manufacturers and different types to be inter-connectable and mutually adaptive to one another, and to realize the digital life for the customers. As long as the user stays anywhere at home, he could access the photos, music and video films via the network. Therefore, as long as the video facilities are conformable to DLNA, they could be directly connected, synchronized and adopted for data transmission to one another without any driver and any interconnecting device. Till the year of 2008, DLNA alliance members have been reaching 245 (vendors), it includes quite a few manufacturing leaders of electronic products such as HP, Intel, Microsoft, IBM, Panasonic, LG, Philips, SONY, Toshiba, Motorola, Nokia, Samsung, etc. Till September of 2008, there are over 3000 kinds of products that have passed DLNA certification. The product specifications defined in “Interoperability Guidelines, Version 1.0” by DLNA in June, 2004 contain the key components as follows: • • DMS (Digital Media Server): It provides the functions of media file access, recording/producing, storing and being used as a source, it is just like a multimedia file server. This type of device includes set-top box, video (DVD) player, PC with built-in media server function, broadcast receiver and HD-embedded home theatre. DMP (Digital Media Player): This device generally represents a device that could search and play or output any media file provided by DMS; these types of devices include TV, printer, home theatre, multimedia cellular, PDA and some specified game terminal machine. “Interoperability Guidelines, Version 1.5” is the recent version published in March of 2006, and, it has been partially expanded in October of the same year. Inside the v1.5 specification, several product specifications have been added in addition to the previous DMS and DMP. It includes: ◦ M-DMS (Mobile Digital Media Server): It is the server defined specifically against the electronic products like handheld mobile phone that 169 Interworking of IP Multimedia Subsystem and Vehicular Communication Gateway ◦ ◦ ◦ is smaller in size, light in weight and portable; the multimedia format it supports is somewhat different from that of the general DMS. M-DMP (Mobile Digital Media Player): It is also designed specifically for the player of mobile device; the multimedia format it supports is somewhat different from that of the general DMP. DMC (Digital Media Controller): When used as a remote device, it could search the multimedia files that could be played on DMS and could assign the DMP either to play that multimedia file or to control the multimedia file uploading or downloading to DMS. In addition to remote device, an intelligent device equipped with basic operation interface could also be used as DMC. DMPr (Digital Media Printer): A DMPr printer could provide printing function under the DLNA network framework. Basically, the DMPr printer function is similar to the traditional USB printer. Concerning the mutual detection among one another facilities, the UPnP standard introduced in previous section is adopted. DMP facility could search the mutually matched DMS facilities on the network via UPnP mechanism, after successful connection, one could proceed with the following playing or data transmission actions. These actions are all automatic, the user needs not to do extra configuration. As to the media content discovery, it adopts also the UPnP mechanism. Presently, it adopts HTTP standard protocol regarding the transmission. RTP will be then added in the future version and be used as the protocol for performing audio and video streaming. In the multimedia formats that DLNA supports for transmission, it is divided again into required 170 Figure 8. DLNA certified logo (Source: DLNA Website) support and optional support. The graphic file format of required support is JPEG, audio format is LPCM, video format is MPEG-2; and, the graphic file formats of optional support are PNG, GIF and TIFF, audio compression formats are AAC, AC-3, MP3 and WMA9, the video compression formats contain MPEG-1, MPEG-4, AVC, WMV9, etc. For product test and certification, DLNA provides standard specification and the software program that allows each vendor to do the selftest. The vendor could do the self-test first and verify its normal execution, then, sent to DLNA certifying organization for further certification. After passing the certification, vendor will be awarded DLNA certification emblem as shown in Figure 8. Every single famous consumer electronic product manufacturer has seen the DLNA vision relatively promising. The currently adopted method for solving the mutual communication among the facilities is also suitable, however, the present DLNA standard specification may be still in some way not enough, such as the Digital Right Management (DRM) of the multimedia file has been in great short, and at the same time, it also lacks security protection mechanism; it could be easily attacked by hackers or virus and thus causes the hidden concerning about the whole DLNA environment. Since DLNA has formulated the key product categories and network protocol framework, this allows the products made by different vendors to have a mutually communicative interface and platform, and solidly makes the vision of digital home/vehicular network be realizable in the coming years of near future. Interworking of IP Multimedia Subsystem and Vehicular Communication Gateway inTerworking archiTecTure and design design of osgi gateway that supports sip In a car, there could exist many kinds of network devices and Networked Appliances (NAs), the communication protocols adopted by different devices are also not all the same; in order to make in-car NAs to be able to inter-communicate with each other and to have a unified interface to administer and control NAs, OSGi Alliance has formulated a open service platform named OSGi Service Platform (abbreviated as OSGi SP) to assist in administering and controlling in-car NAs (such as X.10 device, HAVi device, Jini device, UPnP device, etc.). After having OSGi SP, in order to allow the user to be able to remotely control the in-car NAs, one must make some updating and modifications on the OSGi SP; to allow the user to connect via the network to OSGi Residential Gateway (RGW) even out of car and to further control the in-car NAs (the OSGi SP is practically implemented on the OSGi RGW). Here, we will discuss the technology regarding the user’s utilization of SIP device outside the house to transmit SIP message to remotely control the in-car devices. The applied network framework is shown in Figure 11.9. In order to let the user be able to communicate with each other while using devices of different communication protocols (UPnP or SIP device), except the OSGi Service Bundle that OSGi already supports, one must depend on SIP Service Bundle and Bridging Bundle (Bushmitch et al., 2004; Brown et al., 2006) that have been provided additionally by OSGi platform. OSGi Service Bundle provides the methods of device discovery and communication; the major function of SIP Service Bundle is to enable the device and the bundles to communicate with each other through SIP and to fulfill the functions of device mobility and service mobility, etc. Besides, SIP Service Bundle could be used to bridge two or more than two OSGi RGWs on the cars to make the RGWs to be communicative with each other. After having SIP Service Bundle and Bridging Bundle, an outside user could use his SIP device to transmit SIP message to remotely control the NAs on the car. Firstly, after the outside user sends out the SIP message, the message will be sent to OSGi RGW via the SIP Proxy Server on the Internet, and then gone through the SIP proxy Figure 9. OSGi residential gateway supporting SIP 171 Interworking of IP Multimedia Subsystem and Vehicular Communication Gateway Figure 10. SIP service bundle(Brown et al., 2006) server provided by SIP Service Bundle, the SIP message will be sent to a SIP UA. The SIP UA could probably be the SIP UA built in SIP device, it could also be the virtual SIP UA offered by SIP Service Bundle. The former will directly send the SIP message to that SIP device and let the SIP device process the received messages by itself; the latter will be forwarded to Bridging Bundle, the SIP message will be translated into certain specific message that is able to control the NAs (e.g., UPnP message format). In the following section, we will sequentially introduce the functional components and their operation modes of SIP Service Bundle and Bridging Bundle. SIP Service Bundle As shown in Figure 10, SIP Service Bundle exists in a form of OSGi bundle within the OSGi framework; that responsible for providing SIP support for OSGi Bundles and the device that registers in OSGi is its major function. The features of SIP Service Bundle contain service registration, messaging, event subscription/notification and the complete SIP proxy and SIP server functions. Once the bundle activates, it will make a registration in OSGi service registry to become 172 a SIP service. After the SIP bundle registering, other bundles could then use this SIP service to register to become a SIP device. Then, the outside user could control the in-car SIP device via the SIP message. SIP Service consists mainly of three objects; they are respectively SIPServer, SIPDevice and SIPUserAgent. SIPServer – SIPServer, this object represents the SIP Service server (proxy/registrar) in OSGi RGW. If being served as a proxy server, it will accept the registration message sent from SIP device or bundles, and send the received SIP message to the registered SIP UA. In addition, SIPServer could serve as a registrar server, allowing SIP device or the non-SIP device of SIPUserAgent to register with the registrar, and providing some additional ways for the non-SIP device to be able to register with the registrar server located at the external network. Using SIPServer object could let SIP Service Bundle have the following several functions: 1. Asking a SIPUserAgent to serve as a virtual SIP UA: When there is a non-SIP device that intends to register with a registrar server, the SIP Service Bundle will request a SIPUserAgent to serve as a virtual SIP Interworking of IP Multimedia Subsystem and Vehicular Communication Gateway 2. 3. UA. A virtual SIP UA has all SIP capabilities, but it is not a physical SIP device. Once SIPUserAgent has been produced, it could automatically register with the registrar server. To acquire a SIPDevice object that represents that device for any virtual or physical SIP device that has registered with the OSGi framework: the SIP server must be responsible for doing a service registry on the OSGi framework for those SIP devices that have registered with the SIP server, to register them as the SIPDevice service. After that, the other bundles in OSGi framework could then use those SIP devices. All of the SIP Service Bundle relevant events that happened in OSGi framework will notify with the SIP Service Bundle. For example: the devices registering/unregistering. SIPDevice: SIPDevice, this object could represent a physical SIP device and could also represent a virtual SIP device. A SIP device could register with the registrar to become a physical SIP device; a non-SIP device could register as a virtual SIP device. This virtual SIP device is just a bundle that requests a SIPUserAgent function. When a device has finished the registration, the SIP Service Bundle will produce a SIPDevice object for that device to represent it. This will allow other bundles in the OSGi framework to be able to control the device which the SIPDevice represents. SIPUserAgent: When a non-SIP device intends to register with a registrar server or communicate with other SIP devices, it must support the function of SIP UA. The SIP device itself supports SIP UA, so it could register with registrar server or communicate with other SIP devices directly. However, a non-SIP device (such as air-conditioner, fridge or lamp) must ask SIP Service Bundle for a SIPUserAgent to serve as a virtual SIP UA in order to register with registrar to become a virtual SIP device or to communicate with other SIP devices. Bridging Bundle In order to perform the message translation between two devices that adopt different com- Figure 11.SIP-UPnP bridging scenario (Brown et al., 2006) 173 Interworking of IP Multimedia Subsystem and Vehicular Communication Gateway munication protocols, bridging is a compulsory process. Bridging is a mechanism of applicationlayer proxying. For example, translating the SIP message into UPnP message or translating the UPnP message into the SIP message. The software developer could use the Application Programming Interface (API) provided by OSGi platform to import the required package (such as org.osgi. service.upnp or org.osgi.framework), he could then develop his own Bridging Bundles (such as SIP-UPnP Bridging Bundle). Bridging Bundle is an OSGi bundle. It possesses the functions of asking a SIP UA from the SIP Service Bundle and could also ask the OSGi Service Bundle for an UPnP function. So, the devices that adopt different communication protocols could intercommunicate with each other via different types of Bridging Bundles. One of the bridging examples is to utilize a portable SIP mobile light control device to remotely observe and control the in-car UPnP desk lamp status. Here, we assumed the desk lamps are all UPnP devices. This kind of application example is to elucidate that no matter where the users are, they could remotely control the in-car devices via the network. This SIP lighting controller could acquire the UPnP desk lamp relevant information from the vehicular UPnP network environment via Bridging Bundle. The controller cannot directly use the infrared technology to control the desk lamp; on the contrary, it has to depend on the Bridging Bundle to interrogate the UPnP Service. The controller could also subscribe the status changes with respect to every desk lamp. Therefore, Bridging Bundle will listen the event messages of all in-car desk lamps and utilize the SIP NOTIFY message to forward the received event message to controller. In order to control this UPnP light, controller will send out SIP MESSAGE to Bridging Bundle, and the Bridging Bundle will translate the SIP MESSAGE contents into the function call of UPnP Service API; in the end, it will produce SOAP control message and send them to UPnP light (the desk lamp). 174 Please refer to Figure 11 for the flow chart of this example. This type of bridging is applied on the same OSGi platform; and for the other type of application, it belongs to two devices that adopts different communication protocols; about how to mutually communicate with each other across the different OSGi platforms. In the next section, we will briefly introduce how the bridging operates across the OSGi platforms. inter-gateway bridging In the previous section, we have introduced how to enable the mutual communication between the SIP and UPnP device via an OSGi RGW. Therefore, in this section, what we will discuss is another situation of application. Assuming the user owns some UPnP devices and an OSGi RGW in his house. When he drives his car which equipped with an OSGi RGW to travel, if he wishes to access the data in his home UPnP device, he must rely on the assistance of the SIP Service Bundle and Bridging Bundle embedded on the present mobile RGW on his car and the in-house home RGW in order to translate the SIP and UPnP messages properly. As shown in Figure 12, if a user owns an UPnP DVD player at home, and this DVD video recording/player has registered with the SIP registrar server in the SIP Service Bundle to become a virtual SIP device. Assuming the user wishes to use the UPnP TV on his car to connect back to the DVD video player at his home to watch the film stored in the DVD player. First, the user will operate this UPnP TV, the TV sends out UPnP message (SOAP) to the UPnP Service Bundle. This UPnP Service Bundle is responsible for providing the relevant information of the service and device that exist on every registered UPnP device. UPnP service will pass the message to OSGi service registry via service API. OSGi service registry will then pass the UPnP control message to SIP-UPnP Bridging Bundle. Therefore, UPnP message will be translated into Interworking of IP Multimedia Subsystem and Vehicular Communication Gateway Figure 12. Inter-Gateway bridging example SIP message in SIP-UPnP Bridging Bundle, and, be sent to virtual SIP UA that the UPnP TV has registered on the SIP Service Bundle. In the end, via the SIP Service Bundle on the mobile RGW, the SIP message is sent back to SIP Service Bundle on the home RGW. The processing sequence of this message in the home RGW is contrary to that in the mobile RGW, and the SIP message will be translated into the UPnP message in XML format inside the SIP-UPnP Bridging Bundle of the home RGW, and sent to UPnP DVD player. Therefore, the user could establish the RTP package route between mobile RGW and home RGW, and pass the RTP stream from the UPnP DVD video player located at the home RGW to the UPnP TV located at mobile RGW. This is an example of integrating home network and vehicular network through inter-gateway bridging. sip and upnp mapping UPnP has adopted many standard communication protocols in order to make it compatible with the present network communication protocols; therefore, UPnP could be a cross-platform protocol. Except the often heard protocols – IP, TCP, UDP and HTTP, in the following, some examples will be given regarding the not-often-heard UPnP protocols: SSDP (Simple Service Discovery Protocol) It is used when the UPnP intends to execute the Discovery phase. It is exactly the timing when the control point of the home/vehicular network (exactly the above-mentioned RGW) interrogates 175 Interworking of IP Multimedia Subsystem and Vehicular Communication Gateway the device of its interest about what services are available or when a device is added into the UPnP network, it will automatically broadcast the services it provides to inform the control point or RGW. SOAP (Simple Object Access Protocol) It is used when the UPnP intends to execute the Control phase. Once the control point or RGW has acquired the relevant description of certain specific device, it will be known that what actions could be applied on that device, so, the control point or RGW will use the SOAP to issue commands to control the status of that device. GENA (General Event Notification Architecture) It is used when the UPnP intends to execute the Event phase. The user could subscribe the device it intends to control via the control point or RGW; if the status of that device has changed, the device could use GENA to issue the event notification to the control point or RGW to inform its present status of that device. The mapping way between SIP and UPnP is to utilize SIP message to control UPnP device and the event notification. Therefore, regarding the message interchange method between SIP and UPnP, except the above-mentioned UPnP communication method, it also includes three types of major SIP communication ways; they are respectively SUBSCRIBE, NOTIFY and MESSAGE. In the following, we will introduce how SIP and UPnP could achieve the messages translation. SUBSCRIBE/NOTIFY When an outside user wishes to use the SIP device registered in the RGW to connect back to the OSGi RGW on the car and ask to access some functions provided by the UPnP device of interest and the detailed information. At this moment, the user will 176 use his SIP device to send out SIP SUBSCRIBE message to RGW, and the RGW will use the UPnP device address acquired during the Discovery phase to execute the Description phase by using HTTP (please refer to UPnP introduction). That UPnP device will utilize HTTP to return the detailed information in XML format to OSGi RGW, and the Bridging Bundle in OSGi RGW will then translate this message into SIP NOTIFY message and send it back to the SIP device. In addition, a user could subscribe the device status from OSGi RGW in advance. When the status of that device changes, the OSGi RGW has to inform the SIP device. First, the user will send out SIP SUBSCRIBE message to OSGi RGW via his SIP device, once the status of the UPnP device that we intend to subscribe has changed, the UPnP device will use GENA protocol in XML format to send the event message back to OSGi RGW; then, the Bridging Bundle in OSGi RGW will translate this message into SIP NOTIFY message and send back to the SIP device to inform the status change of that UPnP device. MESSAGE After the SIP device has acquired the detailed description of the UPnP device one wishes to control, one could send out the SIP MESSAGE to OSGi RGW. The Bridging Bundle in OSGi RGW will again translate this SIP message in XML format and use SOAP protocol to control UPnP device. integrating ims and dlna connected home The concept of Connected Home is originated from the common vision of DLNA alliance and UPnP organization. No matter what brand the home devices belong to, they could communicate with each other, share the multimedia data, the home surveillance and remote access service. However, in order to let a user access remotely the device Interworking of IP Multimedia Subsystem and Vehicular Communication Gateway at home/car and enjoy the guarantee of security and quality of service at the same time, it must need a reliable system to achieve the remote access service. In order to achieve this goal, Ericsson has proposed a framework (Fasbender et al., 2008) that has combined the advantages of IMS, DLNA and UPnP as shown in Figure 13. The user does not need to purchase extra specific device; this architecture utilizes the existing IMS framework to provide the authentication and authorization of the user identity, message route, and to establish the secured multimedia sessions and QoS guarantee. Besides, the design of this framework is fully conformable to the present standard of the consumer electronic products for providing multimedia data sharing and service delivering. The core component in this framework is the Home IMS Gateway (HIGA) functional component that Ericsson has continuingly developed on the Residential Gateway (RGW) since 2005. higa (home ims gateway) HIGA is currently formulated by the two organizations of (Home Gateway Initiative, 2007; ETSI TISPAN, 2008). It is a standard of NGN home gateway; it can be regarded as a gateway on the car as well. HIGA is a functional component operating on RGW. It is responsible for collecting the relevant information of the in-house/in-car users, devices and services, and managing the IMS session for the home/ in-car devices that do not support IMS function. With the HIGA, RGW then contains an ISIM card (IMS Subscriber Identity Module Card), it allows the in-car devices to access the services provided by IMS network via the RGW. HIGA will also perform the translation between the protocols adopted by the in-car devices and IMS/SIP. HIGA could use IMS SIM card to register with IMS core network in the secured authentication method. The in-car devices could communicate Figure 13. High-level remote access architecture (Fasbender et al, 2008) 177 Interworking of IP Multimedia Subsystem and Vehicular Communication Gateway with the IMS network through the HIGA which is equipped with the functionalities of Back-to-Back UA (B2BUA) and SIP UA. For example, there is a SIP VoIP phone on the car, this phone could then register with HIGA by using the SIP UA it owns by itself, the B2BUA in HIGA could then translate this SIP message into the IMS specified message and forward to IMS network. In reality, this framework is an extension of UPnP Remote Access standard; making it be able to support IMS Remote Access and allowing the concept of DLNA Connected Home to be extended to the outside of the home. The two major functional modules of UPnP Remote Access framework are Remote Access Transport Agent (RATA) and Remote Access Discovery Agent (RADA). These two kinds of modules will be implemented on the Remote Access Client (RAC) and the Remote Access Server (RAS). RATA is used to establish a secured communication channel between the Remote UE (User Equipment) and the Vehicular Network; RADA is used to synchronize the UPnP device message Figure 14. Functional architecture 178 and inter-changed contents between RAC and Vehicular Network. Regarding all of the devices included in this framework (Remote UE, Residential Gateway with HIGA and DLNA Devices), one could refer to Figure 14 for the block diagram of functional components that are supposed to be equipped. HIGA could be implemented on any in-car device, but practically speaking, implementing directly the HIGA on RGW will be the most simple and of highest feasibility because RGW itself supports Network Address Translation (NAT) and Firewall (FW) functions. If the HIGA function is integrated into RGW, when a user at outside wishes to access the in-car devices, this HIGA will be like an end point of the signaling delivering in the IMS network; and, in the vehicular network, the HIGA will be like an UPnP device that intends to send the message to the in-car device. In the very timing that makes the vehicular network and next generation all-IP network to be closely integrated, not only it could provide diverse IMS multimedia integrated services to the user, it could also allow the user to access any device of his own no Interworking of IP Multimedia Subsystem and Vehicular Communication Gateway matter anywhere, this will make our lives more convenient and digitized. ims approach for remote access With the supports of DLNA and UPnP in the vehicular communication environment, we could establish the remote access session between the Remote UE and the RGW that supports HIGA via the IMS cooperation. Assuming the Remote UE has a digital multimedia renderer (DMR) which supports DLNA, the Network Attached Storage (NAS) supports the function of digital multimedia server (DMS), and both the Remote UE and HIGA have registered with IMS and could connect to IMS network. Therefore, the outside IMS user could connect to the RGW on the car via the IP Multimedia Public Identity (IMPU) of the HIGA. The flow example that uses IMS to proceed with remote access could be divided into six phases as shown in Figure 15. Phase 1: Connection Request Assuming Alice is traveling outside the country, she wishes to access the video film stored in the NAS via her mobile device (in this example, it is the Remote UE); the Remote-Access application on her device will send out the IMS INVITE message to the HIGA on her car. When the INVITE message passes IMS network, the user’s IMS home network will add a P-Asserted-Identity into the INVITE message. At this moment, the HIGA will cross-compare the P-Asserted-Identity and the permitted user identities to verify the Remote UE. The SDP (Session Description Protocol) packed in IMS message will be used to inform the RAS located in the HIGA and the RAC inside the Remote UE of the IP address and the port number of the remote access tunnel. The SDP parameters are also used to negotiate the administration of the virtual private network tunnel (VPN tunnel) and encrypted keys that are established between Remote UE and HIGA. Figure 15.Session setup flow between remote UE, HIGA and NAS (Fasbender et al., 2008) 179 Interworking of IP Multimedia Subsystem and Vehicular Communication Gateway Phase 2: Peer-to-Peer VPN Setup over the IMS Media Plane Once the Alice’s device has completed its authentication and authorization procedures, the device has then established a secured media transmission mechanism with HIGA. The connection method adopted by Ericsson will set up the connection flow according to the IMS standard and will make HIGA as the VPN server. Therefore, after the tunnel has been established, the connection between the Alice’s device and the in-car NAS will be exactly like the UPnP connection at the local terminal. Phase 3: UPnP Discovery The original design of UPnP is to allow the device to be able to mutually communicate with each other in a LAN, if one wishes to extend the UPnP to a WAN, one will surely be confronted with some problems. For example, UPnP is using a messageinterchanging method of multicast to find out the UPnP device. However, these multicast packets will be discarded by the routers located on the Internet. Nevertheless, the remote access method formulated in UPnP Remote Access standard could allow the UPnP RAS filtering device to dig out some relevant messages, and using unicast method to transmit the message of device discovery to the Alice’s device located outside. If adopting UPnP RADA mechanism, it could make the RAC of the Alice’s mobile device to be synchronized with the RAS on HIGA, then, Alice could use her device to access those available multimedia servers and the UPnP services provided by the server. RADA could be used to dynamically inform the Alice device of what new devices have been added on the car. multimedia contents are available, and could choose to playback the mentioned multimedia file by downloading or via streaming method. If one wishes to forward any multimedia data through the VPN connection built in phase 2 in the IMS media plane, one could forward the data by using HTTP (the DLNA default transmission protocol) or RTP (DLNA extra supported transmission protocol). Phase 5: IMS Media Plane QoS Upgrade UPnP does not support QoS management in LAN. But, after integrating the telematics and IP Multimedia Sub-system, one could use the QoS supported by IMS standard to make it be able to support the QoS control and administration between the remote UE (Alice’s device) and HIGA. If it is expected to update some QoS parameters for Alice’s device, the device could send out reINVITE or UPDATE message to the IMS network and HIGA to update the QoS status. Phase 6: Content Playout In the last, the multimedia data chosen by Alice could then be played on her mobile device. With the IMS-based remote access support, it could make up some defects that the original UPnP remote access lacks. Those defects include some security holes, the authentication of remote user and the QoS provision. Therefore, the DLNA/ UPnP applications can be extended to the main system of NGN, it could make the vehicular network and mobile communication network to be more diverse and allow the user to have richer experience in the world of multimedia network. Phase 4: Content Selection conclusion Alice chooses the NAS, using UPnP Content Directory Service (CDS) to browse what the In vehicular network applications, in addition to solving the mutual communication issue among 180 Interworking of IP Multimedia Subsystem and Vehicular Communication Gateway the devices in every single vehicular network, one needs to consider the connectivity of several vehicular networks. This will allow us to make a connection back to the RGW on the car no matter where we are to achieve the goals of remote access control and multimedia data sharing, etc. But, the above applications do not possess enough security authentication mechanism with respect to the vehicular network; besides, it lacks certain degree of QoS guarantee regarding the multimedia contents that the user wants to transmit. So, in recent years, there are some organizations of standardization that are formulating the integration framework between the digital home and IMS. The design of this framework requires totally no modification on the present existing terminal equipments (such as IP devices, SIP devices, Bluetooth devices and UPnP devices), it only requires extra adding of HIGA functional component on RGW, it will enable the in-house/in-car devices to access IMS services via RGW, and it also allows the user to connect to the RGW anywhere via the reliable authentication mechanism provided by IMS to perform a further control on the in-house/in-car devices. Because the most time of our lives are spent either out of home, in the office or on the way to somewhere, so, except to integrate the home/ vehicular network with next generation all-IP network framework, on the other hand, DLNA standard is expected to integrate the home network and telematics of the vehicle industry. We could treat a car or any public transportation as a mobile home network, and could also regard a car as a device for remote accessing to home network; allowing the user to connect back home to access any multimedia data while he is on a vehicle and to administer or control the devices in house. By allowing the user to be able to interact with the home/vehicular network no matter at anywhere, this could then connect more closely our network world together. reFerences Alliance, O. M. (OMA). (2009). Retrieved from http://www.openmobilealliance.org/ Brown, A., Kolberg, M., Bushmitch, D., Lomako, G., & Ma, M. (2006). A SIP-based OSGi Device Communication Service for Mobile Personal Area Networks. In Consumer Communications and Networking Conference(CCNC). Bushmitch, D. Lin, Wanrong., Bieszczad, A., Kaplan, A., Papageorgiou, V., & Pakstas, A. (2004). A SIP-based device communication service for OSGi framework. IEEE Consumer Communications and Networking Conference2006. DLNA. (2009). Retrieved from http://www.dlna. org/home ETSI TISPAN. (2008). Retrieved from http:// www.etsi.org/tispan/ Fasbender, A., Gerdes, M., Hjelm, J., Kvarnström, B., Petersson, J., & Skog, R. (2008). Virtually at home: High-performance access to personal media. Ericsson Review, 2. 3GPP. (2008). IP Multimedia Subsystem (IMS); Stage 2, TS23.228 v8.6.0. 3GPP. (2008). IP Multimedia (IM) Session Handling; IM Call Model; Stage 2. TS 23.218 v8.3.0. Home Gateway Initiative. (2007). Retrieved from http://www.homegatewayinitiative.org/ Open Service Gateway Initiative (OSGi). (2009). Retrieved from http://www.osgi.org Rosen, J., Schulzrinne, H., Camarillo, G., Johnston, A., Peterson, J., Sparks, R., et al. (2002). SIP: Session Initiation Protocol. IETF RFC 3261. Third Generation Partnership Project [3GPP]. (2009). Retrieved from http://www.3gpp.org UPnP Forum. (2009). Retrieved from http://www. upnp.org 181 Section 5 Vehicular Ad Hoc Networks and Delay Tolerant Vehicular Networks 183 Chapter 12 MAC Protocols in Vehicular Ad Hoc Networks Chih-Yung Chang Tamkang University, Taiwan, R.O.C. absTracT With the rapid development of wireless technologies, the Vehicular Ad Hoc Networks (VANETs) have recently received much attention. VANETs technologies aim to ensure traffic safety for drivers, provide comfort for passengers and reduce transportation time and fuel consumption with many potential applications. The achievement of these aims highly relies on efficient MAC protocols which determine the performance of packet transmission in terms of success rate, delay, throughput and bandwidth utilization. This chapter reviews the existing MAC protocols developed for VANETs. Initially, the IEEE 802.11p and DSRC standard are reviewed. Three TDMA-based MAC protocols, called CVIA, VeSOMAC and D*S, are then introduced. In addition, three MAC protocols that cope with the emergency-message broadcasting problem are proposed. Finally, a reliable MAC protocol which is developed based on the cluster topology is reviewed. inTroducTion This chapter reviews the existing MAC protocols developed for VANETs. Initially, the IEEE 802.11p and DSRC standard are reviewed in Section 2. Three TDMA-based MAC protocols, called CVIA, VeSOMAC and D*S, are then introduced in Section 3. The CVIA MAC protocol mainly concerns the fairness DOI: 10.4018/978-1-60566-840-6.ch012 and throughput issues on the demand of Internet access in a multi-hop manner. The VeSOMAC MAC protocol intends to reduce the transmission latency for broadcasting a message over the VANETs. The D*S introduces the data access MAC scheduling protocol which aims to develop a scheduler in Road Side Unit (RSU) so that those upload and download requests can be satisfied as more as possible. In Section 4, three MAC protocols proposed for broadcasting the emergency message are depicted. Copyright © 2010, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global is prohibited. MAC Protocols in Vehicular Ad Hoc Networks These MAC protocols aim at avoiding the packet collision and achieving high success rate. Finally, Section 5 reviews a reliable MAC protocol which is developed based on the cluster topology. The conclusions are finally given in Section 6. Figure 1. Protocol stack relating to OSI model ieee 802.11p and dsrc sTandards DSRC (Dedicated Short Range Communications) (ASTM International E2213-03, 2003) is a well known standard supports both Public Safety and Private operations in roadside to vehicle and vehicle to vehicle communication environments. DSRC standard at 5.9 GHz band is projected to support low-latency wireless data communications between vehicles and from vehicles to roadside units. The DSRC specification is meant to be an extension of the IEEE 802.11 technology into the outdoor high-speed vehicle environment. In fact, the Physical Layer (PHY) of DSRC is adapted from IEEE 802.11a PHY based on Orthogonal Frequency Division Multiplex (OFDM) technology. Moreover, the Multiple Access Control (MAC) layer of DSRC is very similar to the IEEE 802.11 MAC based on the Carrier Sense Multiple Access with Collision Avoidance (CSMA/CA) protocol with some minor modifications. DSRC is meant to be a complement to cellular communications by providing very high data transfer rates in circumstances where minimizing latency in the communication link and isolating relatively small communication zones are important. DSRC protocol layer is developed based on physical, data link and applications layers of traditional OSI model. In the Physical layer, DSRC defines physical parameters for uplink and downlink communication and is mainly working in the 5.9 GHz band (U.S.) or 5.8 GHz band (Japan, Europe). In the data link layer, DSRC defines frame format, frame wrapper and the procedures of MAC and Logical Link Control (LLC). Ap- 184 plication layer includes the fragmentation and defragmentation of data application service and service primitive for a variety of applications, including electronic toll collection, emergency warning system, vehicle safety service, commerce transactions via cars, electronic parking payments, probe data collection and so forth. The standard of DSRC is comprised of IEEE 802.11p and IEEE 1609 family. Figure 1 shows the correspondence between DSRC and OSI models. IEEE 802.11p (Draft 7.0, 2009) is a draft amendment to the IEEE 802.11 standard to add Wireless Access in the Vehicular Environment (WAVE). As an extension of IEEE 802.11, IEEE 802.11p defines how data exchange between high-speed vehicles and between the vehicles and RSU in 5.9 GHz (5.85-5.925 GHz) band. IEEE 1609 is a higher layer standard based on IEEE 802.11p. The IEEE 1609 family is consisted of IEEE 1609.1 (2006), IEEE 1609.2 (2006), IEEE 1609.3 (2007) as well as IEEE 1609.4 (2006). IEEE 1609.1 plays the role of WAVE Resource Manager which specifies a DSRC application overlying WAVE and allows remote site applications to communicate with OBUs or RSUs. As a standard of application layer, IEEE 1609.1 conducts application-level information interchanges. IEEE 1609.2 supports WAVE security services for applications. It defines secure message MAC Protocols in Vehicular Ad Hoc Networks Figure 2. The frequence map of DSRC formats, specifies methods for securing WAVE management messages and application messages and processes secure messages of DSRC/WAVE systems. In addition, it handles exception of vehicle-originating safety messages and provides administrative functions necessary for core security functions. IEEE 1609.3 provides WAVE devices and systems with WAVE networking services. It mainly defines WAVE networking services operated at the network and transport layers. In other words, IEEE 1609.3 represents roughly layers 3 and 4 of the OSI model and the IP, UDP and TCP elements of the Internet model. IEEE 1609.4 supports frequency band coordination and management within the MAC layer, supplements 802.11 MAC features as well as coordinates operations on both the Control Channel (CCH) and Service Channels (SCH). Figure 2 depicts the Frequency band of DSRC. The frequency band used by DSRC ranges from 5.855GHz to 5.925GHz and is divided into seven 10MHz communicating channels and one 5MHz reserving channel. The 5.885~5.895GHz is control and safety emergency channel. The rest of band is reserved for public safety or private operations. Furthermore, two specific channel pairs, channel 174 with channel 176 and channel 180 with channel 182, are able to merge to be a 20MHz channel for special conditions. Tdma-base mac proTocols on VaneTs A number of VANET MAC protocols have been proposed based on TDMA system. Compared with probabilistic-based MAC protocol, an obvious advantage of TDMA system is the ease for avoiding collision and contention with smart assignment for the transmission of each slot. This subsection reviews some TDMA-based MAC protocols developed for VANET. cVia mac protocol for Vehicular networks An efficient TDMA-based MAC protocol can assign each transmission with a slot without collision such that potential parallel transmissions can be exploited. Korkmaz et al. (2006) proposed a new cross-layer protocol, called Controlled Vehicular Internet Access (CVIA), for vehicular Internet access along highways protocol. The proposed CVIA protocol can be applied on the application such as Internet access through RSU in a multihop manner. In the network environment of CVIA, the service region is not fully covered by the deployed RSUs due to hardware cost constraint. As a result, vehicles might need to exchange data with RSU via multi-hop communication. Since vehicles with longer distance to the RSU have fewer opportunities to exchange their data with the gateway, one major problem considered in this research is the fairness. Other issues considered in 185 MAC Protocols in Vehicular Ad Hoc Networks Figure 3. Concepts of slots and segments in CVIA CVIA include collision avoidance and throughput enhancement. As shown in Figure 3, the proposed CVIA divides service region into a number of equalsized segments. Segment Si closer to the RSU will be assigned with a smaller identifier number i. The CVIA protocol assigns a specific time slot to each segment so that vehicles belonging to the segment can access the channel with collision. The objective of CVIA is to increase the end-to-end throughput while achieving fairness in bandwidth usage between road segments. in In each segment, two vehicles TRi and TRiout will be selected as temporary routers for collecting and then forwarding data packets. The TRiin takes charge of forwarding data sent by the vehicular in the neighboring segment with bigger ID whereas TRiout is mainly responsible for collecting the packets to the forward of next Figure 4. Packet movement via routers 186 segment closer to RSU. As shown in Figure 4, the was forwarded by forwarders data sent by TRiout +1 in out TRi , TRi and TRiin-1 subsequently. Three major tasks should be performed by vehicles of some specific segment in the allocated the time slot. Figure 5 gives an example for illustrating the basic concept of CVIA. As shown in Figure 5a, the TRiin ,next delivers the packet train originating from segment Si+1 to TRiout . After, as shown in Figure 5b, local packets of the segments Si are gathered by TRiout in a collision avoidance manner. Finally, the TRiout of segment Si creates a new packet train based on received packets and then sends it to TRiin-1 in Si-1 , as shown Figure 5c. As shown in Figure 6, the CVIA algorithm mainly consists of six phases, namely Inactive, Vehicle Position Update, Temporary Router Selection, Intra-segment Packet Train Movement, Local MAC Protocols in Vehicular Ad Hoc Networks Packet Gathering Phase as well as Inter-segment Packet Train phases. Figure 7 gives an example that would be used throughout this section for illustrating the detail operations of each phase. (a) Inactive Phase. Vehicles in this phase are not permitted to exchange data for avoiding collision and contention. For a given time slot, each vehicle can easily know whether it stays in inactive state or not. The reason is stated below using the example shown in Figure 7. We assume that each RSU serves six segments Si, 1 ≤ I ≤ 6. In addition, we assume that vehicles are able to obtain their positions and synchronize their clocks through the equipped GPS. A set of time slots are divided into two sets, odd and even sets, depending on the number labeled on the slot is odd or even. The six segments are also divided two sets, namely the odd set {S1, S3, S5} and the even set {S2, S4, S6} Segments belonging to the same set will be active at the same time slot. Herein, we assume that the even and odd time slots are allocated to even and odd segment sets, respectively. For example, vehicles in segments S1, S3, S5 are active and are permitted to exchange data in the odd slots. All vehicles can be aware its segment number according to its location information. Therefore, each vehicle knows whether or not it should be active for a given time slot. A vehicle that stays in inactive phase will do nothing. (b) Vehicle Position Update Phase. This phase mainly exchanges the location information of all vehicles located in the active segment. The location information will be further used in the later phases. Each vehicle should first checks if it stays in active state for a given time slot. If it is the case, the vehicle should execute the operations defined in the Vehicle Position Update Phase. Let time interval tu is the length of this phase. In this phase, each vehicle picks a Random Waiting Time (RWT) from 0, tu - tPUP, where tPUP is the time duration required for sending a packet. After an RWT, vehicles access the channel using the Distributed Coordination Function (DCF) as defined in the IEEE 802.11 protocol. (c) Temporary Router Selection Phase. The topology of the vehicular network is dynamic. New temporary routers should be selected periodically, because the router may move to the next Figure 5. Three important phases in CVIA 187 MAC Protocols in Vehicular Ad Hoc Networks Figure 6. CVIA protocol state machine segment. The selected routers are called TRiin ,next and TRiout ,next until they become active. The TRiin is responsible for selecting new TRiin ,next and TRiout ,next after the position update phase. The TRin uses two parameters, router lifetime and safe area, to choose the new routers. The router lifetime represents the time which routers should stay in the segment. Therefore, the router times of TRin and TRout will be different because they have different tasks as described below. The TRiout is responsible for three tasks in the active time. First, TRout receives the packet train sent from TRiin . After, TRout collects the local packets from all vehicles in the same segment. Finally, TRout creates a new packet train and then sends this train to TRi-1 in which is located in the next segment. Hence the router lifetime of TRiout should be long enough for executing the three abovementioned tasks. The router lifetime of TRiin is quite different from that of TRiout . 188 First, TRiin receives the packet train coming from segment Si+1. Then it selects and announces the TRiin ,next and the TRiout ,next at the beginning of the next active slot. Final, TRiin forwards the packet train to TRiout in the next active slot. As a result, the router lifetime of TRiin should be carefully calculated accordingly. The candidates of TRiout ,next and TRiin ,next should exclude the vehicles which will move out the segment during the router lifetime. The farthest and closest candidates away from the RSU will be selected as TRiin ,next and TRiout ,next , respectively. Finally the TRiin will broadcast the result of new inheritors. (d) Intra-segment Packet Train Movement Phase. In this phase, a priority policy should be designed to guarantee that the forwarding packets can be delivered to the Trout of the same segment. The priority of TRiin transmission should be highest since it is in charge of forwarding packets sent MAC Protocols in Vehicular Ad Hoc Networks Figure 7. Slot allocation of CVIA from router of previous segments. To achieve this, TRiin only need to wait for a SIFS period as its active slot starts. (e) Local Packet Gathering Phase. In this phase, TRi-1 aims to receive the data packets of those vehicles in the same segment. All vehicles in segment Si have the same priority for accessing channel. Therefore they access channel in a contention-based manner. To avoid the occurrence of collision, each vehicle should wait for a random period of time. An awaken sender can access channel if it detects that the medium is idle. In case that the medium is busy, the sender will again wait for another random time period. In this out phase, TRi has the highest priority to transmit in this phase. To maintain the fairness among all vehicles in different segments, a predefined size for local packets will be the constraint for data transmission in each segment. Since Trout is responsible for receiving all local packets in this phase, it can be aware whether or not the size of received packet is larger than the predefined 189 MAC Protocols in Vehicular Ad Hoc Networks size. If it is the case, the Trout should broadcast a message to terminate this phase. (f) Inter-Segment Packet Train Movement Phase. In the last phase, TRiout will create a new packet by combining the packets collected in the Intra-segment Packet Train Movement Phase and Local Packet Gathering Phase. The new packet will be sent to the router of next segment. Then this phase will be terminated. All vehicles move back to the inactive state and wait for the next active slot. In summary, the CVIA protocol improves the performance of transmission in the environment which takes into consideration the vehicle-to-vehicle and vehicle-to-roadside communications. The packet collision ratio of CVIA decreases and the fairness among segments are achieved. However, the fairness is achieved between segments, rather than between vehicles. A particular case that one segment has light traffic demands and the other segment has heavy traffic demands, allocating same bandwidth to all segments is not a good way. In addition, the bandwidth utilization of CVIA can be further improved. To avoid packet collision, each segment is assigned with an equal-length active slot. However, the constraints of packet size for different segment are different, which results in unused bandwidth occurred in the segment farer to the RSU. As a result, bandwidth utilization can be improved. Besides, the slot-based approach requires accurate time synchronization which is difficult to be implemented in the real world. Vesomac mac protocol In a VANET, emergency message such as accident information should be broadcasted within an acceptable time period. To avoid the transmission delay raised by collision and contention, a TDMA-based system would be a better candidate than probabilistic-based system. Fan Yu et al. proposed a novel Medium Access Control protocol for inter-vehicular wireless networking using the 190 emerging DSRC standards. A distinctive feature of VeSOMAC (Yu et al., 2007) is its distributed design with fast schedule reconfiguration for coping with vehicular topology changes. The main contribution of the paper is the design of a self configuring TDMA protocol capable of inter-vehicle message delivery with short and deterministic delay bounds. The most significant contribution of this protocol is that it allows a set operations running in a TDMA-based system without strictly clock synchronization. The proposed Vehicular SelfOrganizing MAC (VeSOMAC) assume that each vehicle is aware of its own location and velocity. The following uses an example shown in Figure 8a to illustrate the basic concept of VeSOMAC. Assume that there are four vehicles, A, B, C, and D, move in order. After an emergency event (e.g. an accident) occurred in front of the platoon A-BC-D, the platoon head A periodically broadcasts warning messages instructing other vehicles to slow down for avoiding collisions. Such warning messages are to be forwarded by all vehicles across the entire platoon with minimum possible delivery latency. With an example TDMA allocation with arbitrarily slot placement as shown in Figure 8b, it will take three TDMA frames for delivering the messages to all vehicles in the platoon. However, with a possible VeSOMAC allocation, in which slots are allocated based on the vehicles’ relative locations as shown in Figure 8c, the delivery delay can be significantly reduced. In this example, all messages can be delivered within a single frame. This improvement can be much more pronounced for larger platoons. This way VeSOMAC can effectively enhance highway safety by leveraging its ability to allocate slots based on location, speed, and other vehicular contexts. Information about allocated slots is exchanged among the vehicles using a Bitmap Vector in each packet header. Figure 9 depicts an example to illustrate the concept of bit-map. The slot allocated to vehicle B is marked with blue ink while the slots occupied by all of B’s one-hop neighbors are marked with green ink. Although these neighbors’ MAC Protocols in Vehicular Ad Hoc Networks Figure 8. Basic idea of VeSOMAC slots are shown with respect to B’s frame, each neighbor maintains its own asynchronous frame. The middle of the bitmap vector represents B’s slot time. The bitmap vector is 4-bit long and each bit represents the occupancy status of two slots around B’s own slot. The major reason for using one bit to represent two slots is that the neighbor’s slot can partially overlap with two contiguous slots of B’s frame in the asynchronous mode. For example, the ‘1’ in “+1” location indicates that at least one of the two slots immediately following B’s slot are already fully or partially occupied. Similarly, a ‘0’ in the “-1” location indicates that vehicle B perceives both the slots before its own slot to be free. The bitmap vector length is a design parameter whose maximum value is the frame slot count. In Figure 9, the frame size is 12, whereas the bitmap length is 4, which can convey the occupancy information about only 8 slots. With a bitmap size 4, B is unable to represent the occupancy information about one of its neighbors’ slots-the one in extreme left. Using this header bitmap, a vehicle continuously informs its 1-hop neighbors about the slots occupied by its 1-hop neighbors. By listening to the bitmaps in all received packets, a vehicle can detect the slot locations of its 1-hop and 2-hop neighbors. This information can then be used for choosing a slot which is non-overlapping with the one and two hops vehicles’ slots. Slot allocation in VeSOMAC needs to satisfy the following constraint: 1. 2. Timing constraint: This constraint indicates that no two one-hop or two-hop neighbors’ slots can overlap. Overlaps between onehop or two-hop neighbors cause direct and hidden collisions respectively. Bitmap constraint: For 1-hop neighbors i and j, i’s chosen slot should be able to be represented within the bitmap vector of j. The same is applicable for vehicle j’s slot. In the asynchronous case, since each bit corresponds to two slots, this constraint means that the slots of vehicles i and j can’t 191 MAC Protocols in Vehicular Ad Hoc Networks Figure 9. In-band header bitmap for the asynchronous operation 3. be more than x slots apart, where x is the bitmap length. Ordering constraint: If two vehicles i and j are geographical neighbors and i’s location is ahead of j in the platoon, then i’s chosen slot should be earlier than j’s slot in the time domain. The ordering constraint is optional, and it is useful when the wireless messages flowing from the front to the tail of a platoon are more delay critical than the messages flowing in the reverse direction. Figure 10. State machine of VeSOMAC protocol 192 4. The VeSOMAC protocol state machine with all three constraints is presented in Figure 10. The Stable state for a vehicle indicates the time slot allocation matches the order of geographic location order of vehicles. The Listen and Evaluate are transient states. In the Listen state, vehicle overhears the broadcasting message and look at the bitmap information. Then the vehicle chooses a slot according to its physical location. After a vehicle chooses a slot through the Listen state, it spends a preset (W) number of slots MAC Protocols in Vehicular Ad Hoc Networks in the Evaluate state before getting into the Stable state. In the Evaluate state, the slot is evaluated for W frames to make sure that the vehicle monitors its neighborhood activities to decide if its own allocation became stable. If the assigned slot order is different with the order of geographic locations of the vehicle and its neighbors, it stays in Evaluate state and tries to change its selected slot. When the state machines for all vehicles in a neighborhood reach the Stable state, the protocol is said to have converged. The VeSOMAC proposes a distributed slot allocation mechanism to avoid the packet collision between one-hop and two-hop vehicles, and minimizes the delay of the event delivery by exchanging the Bitmap among neighboring vehicles. Furthermore, the proposed mechanism can be applied in both time synchronous and time asynchronous scenarios. However, the efficiency of the proposed mechanism is highly affected by the frame size, and thus the number of slots in a frame should be carefully determined. If the number of slots in a frame exceeds the vehicles, the chosen slots may be non-continuous and the bandwidth will be wasted. On the contrary, if the slots in a frame are fully occupied, a new vehicle can’t choose an empty slot, which might cause the new vehicle starvation. Therefore, how to decide the frame size by estimating the density of the vehicle network will be an important issue in the future works. data access mac scheduling protocols for rsu Recently, vehicle-roadside data access has received considerable attention in VANET. VANET is a special case in MANET (Mobile Ad-hoc Network) and it is composed of vehicles with wireless communication devices and RSUs. Vehicles can upload and/or download data from RSU. For example, drivers can get lots of traffic or map information from RSU to improve the safety in VANET. Therefore, RSU can be treated as a buffer point (or data island) which can store a variety of data including the value-added advertisement, real-time traffic as well as digital map. As shown in Figure 11, the RSUs are often deployed at the road intersections or areas with high traffic to improve the serving efficiency to each vehicle. In VANET, vehicles often move with high speed. This implies that the time duration for exchanging data between a vehicle and RSU is Figure 11. The scenario considered in RSU scheduler 193 MAC Protocols in Vehicular Ad Hoc Networks Figure 12. Packets in D-list and S-list are sorted by Deadline and Datasize respectively in RSU very short. As shown in Figure 12, when a large number of vehicles intend to exchange data with RSU simultaneously, an efficient scheduler that arranges a sequence of uplink and downlink data access is important. How to decide an efficient serving sequence for satisfying all the services requested from vehicles is an imperative issue. Since each vehicle can only exchanged data with the RSU within a very short time duration which highly depends on the vehicle speed and the communication range of RSU, each request sent from a particular vehicle would have a time constraint. This means that the request would be overdue if the request did not be served within the time constraint. However, the RSU might receive a number of requests with different time constraint. Since the bandwidth resource is limited, how to maximize the network throughput and prevent the requests from being overdue will be the major goal for developing an RSU scheduler. Zhang et al. (2007) proposed a RSU scheduling mechanism that satisfies the services requested from many vehicles under the time and bandwidth constraints. As shown in Figure 12, assume that all of the service requests are 194 buffered in a queue of RSU at the beginning. Each request is characterized by a 4-tuple: < u - id, d - id, op, deadline > , where u - id is the identifier of the vehicle, d - id is the identifier of the requested data item, op is the operation that the vehicle wants to do (upload or download), and deadline is the critical time constraint of the request. In addition to the deadline, the data size should also be taken into consideration because that a transmission for a large sized data needs long service time and may delay other service requests. As shown in Figure 12, the RSU will initially sort all the requests according to the deadline and data size individually and store the sequences as the D-list and S-list, respectively. A weighted DS_value, as defined in Equation 1, is used for RSU to decide the services scheduling sequence. DS _ value = (Deadline - CurrentClock ) ´ DataSize (1) The (Deadline - CurrentClock ) represents the remaining service time. A service with a smaller MAC Protocols in Vehicular Ad Hoc Networks time constraint will have a higher priority while a service with a smaller DataSize will have a higher priority. Consequently, the DS scheme always serves the requests with minimum DS _ value in each scheduling time. emergency mac proTocols on VaneTs broadcast strom mitigation Techniques in VaneTs In VANETs, Broadcasting for emergency message has been widely discussed for a variety of applications. In the application of emergent message dissemination, the packet broadcasting is critical in time constraint. However, packet broadcasting might raise contention and collision phenomenon which might block the dissemination of emergency message. How to develop an efficient broadcasting mechanism is an important research issue in VANETs. Developing emergency MAC protocols can help dissemination of emergency messages with a short propagation delay without packet collision and contention. Wisitpongphan et al. (2007) assumed that each vehicle is aware of location information and proposed three different broadcasting approaches. In the first approach, a node j receiving a packet from node i will check the packet ID and rebroadcasts with probability pij if the packet is received first time. Otherwise, it discards the packet. Let Dij denote the distance between nodes i and j, and R denote the average transmission range. The forwarding probability, pij, can be calculated on a per packet basis using the following simple Equation. The basic concept of Equation 2 is that a receiver closer to the sender will have a larger overlapped communication range with sender and thus has smaller contribution for forwarding the packet to those receivers that have not yet received the same message from the sender. As a result, a receiver that farer to the sender will have a higher probability for rebroadcasting the received message. pij = Dij R (2) To avoid the occurrence of contention, the receiver with small probability value should additionally wait for a predefined time duration. In case of receiving duplicate packets from multiple sources within the waiting period, it calculates a probability for each received message and then selects the smallest probability value as its re-forwarding probability. If a node decides not to rebroadcast, it should buffer the message for another time duration WAITTIME + d , where d is the one-hop transmission and propagation delay, which is typically less than WAITTIME as shown in Figure 13. The proposed approach prevents the packet dissemination from collision. However, the first approach exhibits two drawbacks as described below. First, the vehicle with low rebroadcast priority may still broadcast the packets. Second, Figure 13. Vehicles calculate probability and broadcast 195 MAC Protocols in Vehicular Ad Hoc Networks if the density of VANET is high, several vehicles may have the same probability. They will rebroadcast the packets at the same time, resulting in collisions. The second approach, based on time-slot calculation, is proposed to improve the aforementioned drawbacks. Upon receiving a packet, each receiver checks the packet ID and rebroadcasts with probability 1 at the assigned time slot TSij if the packet is received first time. Otherwise, it discards the packet. The following discusses how to derive the value of the allocated transmission slot TSij. Let the average transmission range and the predetermined number of slots are R and Ns, respectively. Given the relative distance Dij between nodes i and j, TSij can be calculated by Equation 4, where τ is the estimated one-hop delay, which includes the medium access delay and propagation delay, and Sij is the assigned slot number. éN - 1ù ´WAIT + dms êë 8 úû TIME (3) TS = Sij ´ t (4) ij The time slot approach follows the same concept of the weighted p-persistence scheme, but instead of calculating the re-forwarding probability, each node uses the GPS information to calculate the waiting time to retransmit. For example, in Figure 13 the broadcast coverage is spatially divided into four regions. According to Equation 4, a shorter waiting time will be assigned to the nodes located in the farthest region. When a node receives duplicate packets from more than one sender, it takes on the smallest Dij value. Similar to the p-persistence scheme, this approach requires transmission range information in order to agree on a certain value of slot size or number of slots. Note that N 8 is a designed parameter that should be carefully chosen. Although N 8 should theoretically be a function of the traffic density (i.e., the denser the traffic, the smaller the slot 196 size and the larger the number of slots), it is very difficult for each vehicle to predict what the traffic density is and to arrive at a single value of ( ) é min RSS , RSSij - RSS min ) ´ N 8 ùú ê range ( ú Sij = N 8 - ê ê ú RSSrange êë úû in practice. Hence, network designers can, at best, fix this value or adaptively change this value over time; for example the protocol should use five slots during morning and evening rush hour, and three slots during non-rush hours. The second approach might assign the same slot for two vehicles that have similar distances far from the sender. As a result, a collision might be occurred. Another approach is proposed to improve this drawback. The proposed third approach takes into account the previously discussed two approaches. That is, probability-based and slot-based mechanisms are adopted in this approach. Each node in this scheme should also buffer the message for a certain period of time (e.g., éëêN 8 - 1ùûú ´WAITTIME + dms ) and retransmits with probability 1 if nobody in the neighborhood rebroadcasts in order to prevent the message’s dying out. The abovementioned approaches mainly developed based on the location information which is obtained from GPS equipment. However, vehicles may not be able to receive GPS signals in some areas (e.g., tunnels, shadowed areas, urban areas with many high-rise buildings). These broadcasting approaches can also be modified by using the packet Received Signal Strength (RSS) information instead of GPS information. Equations 5 and 6 depict the modifications of approaches 1 and 2, where RSSrange and RSSmin denote the signal range and the value of minimum signal strength, respectively, and the RSSij denotes the signal strength between nodes i and j. pij = RSSij - RSS min RSS range (5) MAC Protocols in Vehicular Ad Hoc Networks ( ) é min RSS , RSSij - RSS min ) ´ N 8 ùú ê range ( ú Sij = N 8 - ê ê ú RSSrange úû ëê (6) broadcast methods for interVehicle communications system Some other studies devoted themselves to develop broadcast schemes for emergency events such as traffic accidents and ambulance alarming. The accidents information should be broadcasted to all vehicles to prevent them from another accident. Therefore, broadcasting mechanism for emergency information is important and has received much attention recently. How to efficiently broadcast the emergency message to all vehicles neighboring to the accident location will be the key technology for preventing secondary accidents. Fukuhara et al. (2005) proposed a novel broadcast method for a vehicular network to timely notify appreciate drivers. They considered that the events can be categorized into two types, according to the broadcasting directions. One type of events is backward-dissemination event which should be broadcast to those vehicles located at the backward location of the accident. A typical example is the traffic collision event. On the other hand, the forward-dissemination event should be broadcasted to those vehicles located at the forward location of the event. For example, the event message of ambulance alarming should be disseminated to the vehicles in the forward direction. Therefore, the broadcast areas of these two types of events should be different. Based on the concept, the authors proposed a broadcast scheme to notify the vehicles in a particular area. Figure 14 shows an example for an ambulance alarming event. The authors define three different areas in the road. The first area is called location offset, which avoids the inaccurate location obtained from a GPS. The second area is used for relaying the emergency message along the forward direction. Vehicles in this area should not only receive the emergency notice but also deliver the notice to other vehicles. Finally, the authors define the area which should be notified. In this area, the vehicles within a proper range can get the notification by other vehicles without forwarding. The following uses an example shown in Figure 15 to discuss how to identify the three areas. Figure 15 depicts the area which consists of an available relay range and an available notifi cation range. Let X and V be vehicle’s location and velocity vectors, respectively. Equation 7(a) depicts the relative direction between the vehicle and the ambulance, where X 1 and X 2 denote the location vectors of vehicle and ambulance, Figure 14. The relay and notification zone 197 MAC Protocols in Vehicular Ad Hoc Networks Figure 15. An example for calculating available relay range respectively, and V 1 and V 2 denote the velocity vectors of vehicle and ambulance respectively. A positive value of Equation 7(a) indicates that the vehicle and ambulance have the same direction. Vehicles that satisfy the criteria of Equation 7(b) should be notified and further relay the emergency message while vehicles satisfy Equation 7(c) but not satisfy Equation 7(b) should only be notified. The calculation procedure in the traffic accidents scenario is similar except the direction is opposite to the ambulance alarming event. X 2 - X 1 + aV 1 ·V 2 > 0 ( p¬ ) (7a) dis tan ce (m.SenderPosition, localPosition ) m.SenderCommunicationRadius (7b) X 2 - X 1 + aV 1 < RNotification (Threshold) (7c) Towards lightweight information dissemination in inter-Vehicular networks In addition to the traffic accidents and ambulance alarming handling, some other emergency protocols are developed for general-type emergency event (Sormani et al., 2006). Davide Sormani et 198 al. intend to reduce the dissemination packets in a broadcasting operation while achieves high success rate. In their design, the packet only consists of the position of the accident, sender’s position, communication range and the propagation function including the target zone. In particular, the propagation function encodes the information about target areas and preferred routes. Figure 16 shows a propagation function associated with a target zone that is reached by a single major road. The function drives messages along the main road — the red line below the function — and towards the target zone — the black ellipse. It is worth to note that the message originator does not compute a predefined trajectory using the propagation function before sending the message. Rather, the route to the destination is the result of the evaluation of the function at each routing hop. For example, a message that is routed outside the black line in Figure 16 does not have to be routed back towards the line, but it can continue its route along a new trajectory, which still ends up on the desired target area. The values of the propagation function can be viewed as the values of the potential to achieve the destination region. Messages should be attracted in the right direction towards decreasing values of the propagation function that is towards areas of minimum potential. Several protocols with the propagation MAC Protocols in Vehicular Ad Hoc Networks Figure 16. Three-dimension diagram of the propagation function function are presented for broadcasting the emergency message with lightweight traffic. a. One Zero Flooding (OZF). The One Zero Flooding (OZF) mechanism is proposed for flooding a message with lightweight traffic by using propagation function. The basic concept of OZF mechanism is that packets would be transferred by vehicles with the smaller value of propagation function. A simple example to illustrate the idea of OZF is shown in Figure 17. In Figure 17, we assume that vehicles A and C both are located within vehicle B’s communication range. We assume that vehicle A and vehicle C calculate the function value between themselves and vehicle B separately after getting a packet broadcasted from vehicle B. If the distance between receiver and target area is shorter than that between sender and target area, the function value calculated by the receiver is smaller than that calculated by the sender. That is, vehicle C has the smaller function value than vehicle B and vehicle A has the larger function value than vehicle B. As a result, vehicle C in the OZF protocol gets the value “one” to relay the packet and vehicle A gets the value “zero” which indicates it does not need to forward the received packet. The OZF mechanism can reduce the phenomenon of packet dissemination along the opposite direction toward the destination. However, several vehicles might forward the received message at the same time if they are within the transmission range of a sender and are closed to the destination. This might raise the contention and collision problem. b. Distance-Driven Probabilistic Diffusion (DDPD). DDPD is proposed for reducing the collision and contention phenomenon occurred in OZF. The DDPD propagation protocol adds a probabilistic choice to OZF. The probability function introduced in DDPD design aims to distribute those transmissions which might be collided with each other when applying OZF. Equation 8 takes into account the distance between the receivers and the sender. Similar to the concept of OZF, vehicles do not forward the emergent messages if these vehicles have larger values of propagation function than sender. On the contrary, vehicles with smaller values of propagation function than sender will forward the emergent messages to the target area with a probability evaluated according to Equation 8. Figure 17. OZF propagation protocol 199 MAC Protocols in Vehicular Ad Hoc Networks FF (X ) = dist (X ,Y ) + Du * BB _ REFR R (8) Figure 18 depicts an example of DDPD. Assume that vehicles B and C have smaller function values than vehicle A and intend to forward the received message with a probability calculated by Equation 8. Herein, we assume that vehicle C has a higher probability value than vehicle B and will forward the received emergency message. As a result, vehicle B will not further forward the message again after it overhears the message broadcasted from vehicular C. Although the DDPD mechanism might reduce the probability of transmission collision, it still has the problem of redundant transmission. For example, in Figure 19, if vehicles B and C are closed to the edge of the communication range of vehicle A, vehicle B will have a high probability for forwarding the packet sent from vehicle A. However, vehicle C would be a better candidate for broadcasting the message than B since it closer to the target area. c. Function-Driven Probabilistic Diffusion (FDPD). Though the abovementioned DDPD mechanism tries to reduce the number of forwarders, however, it does not consider the distance from forwarding candidates to destination in the probability design. To further reduce the number of redundant transmission, FDPD propagation protocol is proposed. The probability function proposed in FDPD relies on the function value by preventing the packet from forwarding in the wrong direction. Different to DDPD, an enhance probabilistic Equation is proposed as shown in Equation 9, where the BEST point denotes the physical location with the lowest value of propagation function within the communication radius of the sender. In the FDPD propagation protocol, vehicles forward the emergent packet if they are located at the right position away from Figure 18. DDPD propagation protocol Figure 19. Vehicles B and C have the same high probability for forwarding the packet sent from vehicle A. However, vehicle C would be a better candidate 200 MAC Protocols in Vehicular Ad Hoc Networks the BestPoint and they are closer to the destination than sender. p¬ f (m.SenderPosition ) - f (localPosition ) f (m.SenderPosition ) - f (BestPo int) (9) As shown in Figure 20, vehicle B and vehicle C intend to forward the received packet with probability 0.75 and 0.99, respectively, according to Equation 9. The key improvement of FDPD is the consideration of BEST point which leads the packet to be forwarded in the right direction. Though FDPD significantly reduce the number of redundant transmissions, however, it does not take into consideration the hole problem, which might occur in a sparse network environment. d. Feedback-Augmented Store and Forward Diffusion (FSFD). Compared with the FDPD, the FSFD further concerns the hole problem which might occur in the light-traffic environment. The FSFD mainly adopts the well known store-andforward mechanism where vehicles can store the emergent packet and move for a period of time, seeking for the possible forwarder. As shown in Figure 21, assume that vehicle B can not find any forward in its transmission range at the time t after receiving the packet from vehicle A. Thus, vehicle B has to store the packet in its queue temporally until it finds forwarder C at the time t + 2. In the meanwhile, vehicle B forwards the packet to vehicle C and therefore the packet will not be blocked at vehicle B. e. Direction-aware Function Driven Feedbackaugment Store & Forward Diffusion (DFDFSFD) Although the FSFD can store and forward the packet for overcoming the hole problem, however, the transmission direction of packet is not taken into account. This protocol takes into account the right direction to target area like the BEST point and applies the FSFD propagation protocol to forward the emergent packet. As shown in Figure Figure 20. FDPD propagation protocol Figure 21. FSFD propagation protocol 201 MAC Protocols in Vehicular Ad Hoc Networks Figure 22. DFD-FSFD propagation protocol 22, the vehicles A and B calculate the angles α and p implies β, respectively. The angle α less than 2 that the packet can be forwarded by vehicle A to the right direction to the target area. On the p indicates contrary, the angle β larger than 2 that the packet forwarded by vehicle B is along the wrong way. reliable mac proTocols on VaneTs Some other studies took the reliable communication into consideration and aimed to construct a reliable network topology for data dissemination. However, in VANETs, the locations of vehicles change with time, which might cause an existed link broken. It is an important issue to construct a reliable backbone topology so that the message flooding can be achieved by the forwarding of those nodes in the backbone topology. This subsection introduces a novel reliable MAC protocol which efficiently constructs a backbone topology for a given VANET. Bononi et al. (2007) proposed a fast and efficient broadcast protocol for VANET. The proposed protocol aims to reduce the influence of vehicles with high speed on performance. The following uses an example to illustrate this protocol. As shown in Figure 23, a distributed clustering dynamic algorithm is proposed to create a dynamic virtual backbone in the vehicular network. The vehicles update their state dynamically. If the vehicles are in the backbone, their states are set to be Backbone Member (BM), otherwise their states are set to be Normal Vehicle (NV). The vehicles in BM are responsible for relaying the massages, while the vehicles in the NV state just need to receive the messages from the vehicles in the BM. The proposed algorithm can reduce the overhead of inter-vehicular communication. Figure 23. The vehicles in BM are responsible for relaying the massages, while the vehicles in the NV state just need to receive the messages from the vehicles in the BM 202 MAC Protocols in Vehicular Ad Hoc Networks A backbone creation process starts whenever a vehicle does not receive backbone beacons for a predefined time interval. As shown in Figure 24, vehicles A, B, C and D initially implement the random back-off mechanism without receiving a message from backbone. In this example, we assume that vehicle A wakes up earliest and sets itself as a backbone member. Then vehicle A chooses its next hop backbone member from neighbors for constructing a robust backbone, the next hop should be able to stay in the communication range of A for the longest time. As a result, the backbone member can be selected subsequently in a hop-by-hop manner. Each vehicle in the BM state should record at most two backbone members which are called previous hop and next hop, respectively. Figure 25 shows that vehicles B and C are the previous hop and next hop of vehicle A, respectively. An important factor for a backbone member to select the next backbone member will be their connection duration. The following Equation is used to calculate the residual communication time between vehicles X and Y. Let RT denote the Residual Time, R denotes the communication range of sender vehicle, and dist(X, Y) denotes the current estimated distance of vehicles X and Y. Let Du = uY - uX denotes the relative speed between vehicles X and Y, where uX and uY are the average speed of vehicles X and Y, respectively. In Equation 10 gives the estimation of residual time of vehicles X and Y, where sign() denotes the function which returns 1 if Du is positive and -1 otherwise. ( ) é max 0, sign Du ù R - dist X ,Y ( ) úû ( ) ê RT (X ,Y ) = ë Du (10) In order to reduce the communication overhead required for constructing a backbone, the number of vehicle in BM state should be minimized. To achieve this goal, Equation 11 calculates the distance of X and Y after a time duration BB_PEFR which is the time required for constructing the next backbone topology, normalized with respect to R. FF (X ) = dist (X ,Y ) + Du * BB _ REFR R (11) If the distance between two vehicles increases, the value of FF would raise. The value of FF would be used as the random backoff parameter after all neighbors of vehicle A calculating RT. Therefore, when a vehicle has the largest value it will wake up first and response the beacon Figure 24. Vehicle A, B, C and D do not receive backbone beacons for a time interval. After random backoff time period, if the first waked vehicle is A, vehicle A will elect itself as a backbone member and choose its next hop. 203 MAC Protocols in Vehicular Ad Hoc Networks Figure 25. Each vehicle in the BM state has at most two neighbors that called previous hop and next hop respectively. Previous hop and next hop are also backbone members. of vehicle A. We assume that vehicle C wake up earliest and it responds a packet to A. Upon receiving the response, Vehicle A records that C is its next hop, and then sends an ACK_WINNER packet to vehicle C. Similarly, vehicle C records that vehicle A is its previous hop and changes the NV state to BM state. All other vehicles give up the contention accordingly. conclusion Vehicular communication systems that can transmit and receive information to and from individual vehicles have the potential to significantly increase the safety of vehicular transportation, improve traffic flow on congested roads, and decrease the number of people of deaths and injuries in vehicular collisions effectively. This chapter review import MAC protocols for VANETs. A number of MAC protocols, including IEEE 802.11p, DSRC standard, TDMA-based MAC protocols, data access MAC scheduling protocol, emergency message and reliable MAC protocols, are reviewed. In this chapter, the basic concept, strategies, advantages and contributions of the reviewed papers are introduced and discussed. We believe the reviewed MAC protocols will be the base for further improvement of protocols in terms of the reliability, scalability, and efficiency in VANETs. The development of efficient VANET protocols that integrate MAC and network layers and the implementation of 204 protocols that meet the requirement of real applications in VANETs will be the future works. reFerences ASTM International E2213-03 (2003). Standard Specification for Telecommunications and Information Exchange Between Road-side and Vehicle Systems 5 GHz Band Dedicated Short Range Communications (DSRC) Medium Access Control (MAC) and Physical Layer (PHY) Specifications. Bononi, L., & Felice, M. D. (2007). A Cross Layered MAC and Clustering Scheme for Efficient Broadcast in VANETs. In Proceedings of the IEEE International Conference on Mobile Ad hoc and Sensor Systems. Fukuhara, T., Warabino, T., Ohseki, T., Saito, K., Sugiyama, K., Nishida, T., & Eguchi, K. (2005). Broadcast Methods for Inter-Vehicle Communications System. In Proceedings of the IEEE Wireless Communications and Networking Conference. IEEEP802.11p (2009, May). IEEE Draft Standard for Information Technology – Telecommunications and information exchange between systems – Local and metropolitan area networks – Specific requirements – Part 11: Wireless LAN Medium Access Control (MAC) and Physical Layer (PHY) specifications Amendment 7: Wireless Access in Vehicular Environments. MAC Protocols in Vehicular Ad Hoc Networks Draft 7.0, IEEE Standard 1609.1 (2006). IEEE Trial-Use Standard for Wireless Access in Vehicular Environments (WAVE) – Resource Manager. IEEE Standard 1609.3 (2007). IEEE Trial-Use Standard for Wireless Access in Vehicular Environments (WAVE) – Networking Services. Korkmaz, G., Ekici, E., & Ozguner, F. (2006). A Cross-Layer Multihop Data Delivery Protocol with Fairness Guarantees for Vehicular Networks. IEEE Transactions on Vehicular Technology, 55(3), 865–875. doi:10.1109/TVT.2006.873838 Wisitpongphan, N., Tonguz, O. K., Parikh, J. S., Mudalige, P., Bai, F., & Sadekar, V. (2007). Broadcast Strom Mitigation Techniques in Vehicular Ad Hoc Networks. IEEE Wireless Communications, 14(6), 84–94. doi:10.1109/MWC.2007.4407231 Sormani, D., Turconi, G., Costa, P., Frey, D., Migliavacca, M., & Mottola, L. (2006). Towards Lightweight Information Dissemination in Inter-Vehicular Networks. In Proceedings of the ACM International Workshop on Vehicular Inter-Networking. Yu, F., & Biswas, S. (2007). Self-Configuring TDMA Protocols for Enhancing Vehicle Safety with DSRC Based Vehicle-to-Vehicle Communications. IEEE Journal on Selected Areas in Communications, 25(8), 1526–1537. doi:10.1109/ JSAC.2007.071004 IEEEStandard1609.2 (2006). IEEE Trial-Use Standard for Wireless Access in Vehicular Environments (WAVE) – Security Services for Applications and Management Messages. Zhang, Y., Zhao, J., & Cao, G. (2007). On Scheduling Vehicle-Roadside Data Access. In Proceedings of the ACM International Workshop on Vehicular Inter-Networking. IEEE Standard 1609.4 (2006). IEEE Trial-Use Standard for Wireless Access in Vehicular Environments (WAVE) – Multi-channel Operation. 205 206 Chapter 13 Routing Protocols in Vehicular Ad Hoc Networks Yuh-Shyan Chen National Taipei University, Taipei, Taiwan, R.O.C. Yun-Wei Lin National Taipei University, Taipei, Taiwan, R.O.C. absTracT Vehicular Ad hoc Network (VANET), a subclass of mobile ad hoc networks (MANETs), is a promising approach for the intelligent transportation system (ITS). The design of routing protocols in VANETs is important and necessary issue for support the smart ITS. The key difference of VANET and MANET is the special mobility pattern and rapidly changeable topology. It is not effectively applied the existing routing protocols of MANETs into VANETs. In this chapter, we mainly survey new routing results in VANET. The authors introduce unicast protocol, multicast protocol, geocast protocol, mobicast protocol, and broadcast protocol. It is observed that carry-and-forward is the new and key consideration for designing all routing protocols in VANETs. With the consideration of multi-hop forwarding and carryand-forward techniques, min-delay and delay-bounded routing protocols for VANETs are discussed in VANETs. Besides, the temporary network fragmentation problem and the broadcast storm problem are further considered for designing routing protocols in VANETs. The temporary network fragmentation problem caused by rapidly changeable topology influence on the performance of data transmissions. The broadcast storm problem seriously affects the successful rate of message delivery in VANETs. The key challenge is to overcome these problems to provide routing protocols with the low communication delay, the low communication overhead, and the low time complexity. inTroducTion The growth of the increased number of vehicles are equipped with wireless transceivers to comDOI: 10.4018/978-1-60566-840-6.ch013 municate with other vehicles to form a special class of wireless networks, known as vehicular ad hoc networks or VANETs (Saha et al., 2004; Xu et al., 2004; Yousefi et al., 2004). Vehicular Ad hoc Network (VANET) is a promising approach for the intelligent transportation system (ITS) Copyright © 2010, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global is prohibited. Routing Protocols in Vehicular Ad Hoc Networks (ASTM E2213-03, 2003). Vehicle information is delivered via the multi-hop wireless transmission over VANETs to provide safety or comfort applications for drivers. VANETs are expected to improve the traffic quality and provide the more convenient driving environment for the general populace. It is known that VANET (Saha et al., 2004; Xu et al., 2004; Yousefi et al., 2004) is a subclass of mobile ad hoc networks (MANETs) (Briesemeister et al., 2000) . Just like a MANET, a VANET has no the fixed infrastructure. In addition, vehicles with high speed mobility make VANETs having quite different characteristics from MANETs; such as rapidly changed topology and frequently network fragmentation. VANETs are mainly realized in city and highway environments. Roads and streets with intersections is the major scenario in the city environment. Multiple lanes with single or dual direction are investigated in the highway environment. These two environments have different impact on VANETs. In the city environment, packets are difficult to be successfully transmitted since the signals are easily shielded by buildings. With the obstacles, two vehicles are not able to be communicated even if they are very close. In the highway environment, the temporary network fragmentation problem is the key issue. Furthermore, VANETs have distinctive features. For example, power constraint is not the major concern, and location information is easily obtained from GPS (Global Positioning System) (Gerten et al., 2005) which is the common equipment in a vehicle. The interest problem is how to develop the efficient routing protocols in VANETs with the consideration of distinctive features of VANETs. The design of routing protocols of VANETs is the important issue for supporting the smart ITS. To enhance the safety of drivers and provide the comfortable driving environment, messages for different purposes need to be sent to vehicles through the inter-vehicle communications. According to the number of receiving vehicles within a geographic region, the roles of destinations are divided into three cases: (1) a static destination or a mobile destined vehicle, (2) more than one vehicle in a geographic region, and (3) all vehicles within a geographic region. When the message should be sent to a static destination or a mobile destined vehicle, unicast routing protocol is utilized. Unicast routing is a fundamental operation for vehicle to construct a source-to-destination path in a VANET. Example of unicast routing is given in Figure 1(a). For the case of the static destination, the routing path is quite different because the source vehicle is continuously moving at the different time period. It is more difficult to continuously trace the location of destination vehicle if the destination is a mobile vehicle. Moreover, when the information is delivered to more than one vehicle in a geographic region, multicast and geocast routing protocols are performed. For the geocast routing, if a vehicle receives a geocast packet from neighbors, the packet should be forwarded or dropped depended on its current location. If this vehicle is located in the specific geographic region, the geocast packet is forwarded; otherwise, the packet is dropped. Multicast in a VANET is defined by delivering multicast packets from a single source vehicle to all members of multicast in a multi-hop communications as shown in Figure 1(b). The multicast and geocast routing protocol also are important functions for many useful applications, including collision warning system, distributed games, replicated file systems, and teleconferencing. Broadcast protocol is utilized if the information should be sent to all vehicles in a network. A source vehicle sends broadcast message to all other vehicles in the network. Example is illustrated in Figure 1(c). Broadcast is an important function in many applications of VANETs, such as advertisement publicity, cooperative operations, group discussions, and route discovery. The design issue of broadcasting is how to effectively prevent packet collision and reduce the broadcasting overhead during the broadcasting. 207 Routing Protocols in Vehicular Ad Hoc Networks Figure 1. Applications for VANET (a) unicast routing, (b) geocast and multicast routing (c) broadcast routing Many results (Safa et al., 2009; Sawamura et al., 2008; Manoharan et al., 2008; Hughes et al., 2006; Manoharan et al., 2005) on MANETs have been proposed for unicast, multicast and geocast, and broadcast protocols. However, VANETs are fundamentally different to MANETs, such as the property of mobility and rapid changed topology (Blum et al., 2004); therefore, VANET is subject to frequently network fragmentation and a stable routing path is not easily to establish. This key differentiation causes the existing routing protocol on MANETs can not be directly applied to VANETs. In this investigation, the recent new results for VANET routing mechanism are first surveyed. As shown in Figure 2, the survey is structured into three broad categories; unicast, multicast and geocast, and broadcast approaches. It is observed that carry-and-forward is the new and key consideration for designing all routing protocols in VANETs. The key ideas of representative technologies in each category are described. In the unicast approaches, existing unicast routing protocols are classified into min-delay and delaybounded routing protocols. The min-delay routing protocols (Lochert et al., 2005; Zhao et al., 2006; Naumov et al., 2007; Chen et al., 2009; Taleb et al., 2007; Sun et al., 2006; Wan et al., 2008) are to construct a routing path with the minimum delay time for a source-to-destination path. The delay- 208 bounded routing protocol (Skordylis et al., 2008) is to establish a minimum communication overhead routing path under a constrained delay time. For the min-delay routing protocol, we review GPCR (Lochert et al., 2005), VADD (Zhao et al., 2006), CAR (Naumov et al., 2007), DIR (Chen et al., 2009), ROMSGP (Taleb et al., 2007), reliable routing (Wan et al., 2008), and GVGRID (Sun et al., 2006). For the delay-bounded routing protocol, we describe delay-bounded routing in (Skordylis et al., 2008). In multicast and geocast approaches, two categories of protocol are discussed, spatial routing protocol and spatiotemporary routing protocol. For spatial routing protocol, DRG (Joshi et al., 2007) and IVG (Bachir et al., 2003) are discussed. For spatiotemporary routing protocol, a mobicast routing protocol (Chen et al., 2009) is introduced. Finally, the broadcast approaches are investigated. The impact of broadcast storm problem on MANETs has been firstly defined in (Tseng et al., 2002), then some researchers (Tonguz et al., 2006) further discusses the impact of broadcast storm problem on VANETs. Besides, two broadcast approaches for VANETs, DV-CAST (Tonguz et al., 2007) and broadcast methods for V2V communications (Fukuhara et al., 2005), are finally reviewed. The challenges and perspectives of routing protocols for VANETs are finally discussed. Routing Protocols in Vehicular Ad Hoc Networks The remainder of this paper is organized as follows. Section 13.2 reviews unicast routing protocols in VANETs. Section 13.3 introduces multicast and geocast routing protocols in VANETs. Section 13.4 describes broadcast routing protocols in VANETs. Section 13.5 concludes this chapter. unicasT rouTing proTocol This section introduces the unicast routing protocols in VANETs. The main goal of unicast routing in VANETs is to transmit data from a single source to a single destination via wireless multi-hop transmission or carry-and-forward techniques. In the wireless multi-hop transmission, or called as multi-hop forwarding, technique, the intermediate vehicles in a routing path should relay data as soon as possible from source to destination. In the carry-and-forward technique, source vehicle carries data as long as possible to reduce the number of data packets in the VANETs. The delivery delaytime cost by using carry-and-forward technique is normally longer than using wireless multi-hop transmission technique. Two categories of routing protocol designing are classified, min-delay routing protocol and delay-bounded routing protocol. Min-delay routing protocol aims to minimize the delivery delay-time from source to destination. Delay-bounded routing protocol attempts to maintain a low degree of channel utilization within the constrained delivery delay-time. Figure 3 gives the classification of these unicast protocols. min-delay routing protocol The goal of min-delay routing protocols is to transmit data packets to destination as soon as possible. The transmission delay time is the major concern and the shortest routing path is usually adopted. However, the shortest routing path may be not the quickest path with the minimum delay time in VANETs. The shortest routing path may be found in a low density area, packets can not transmit by the multi-hop forwarding since that there is no neighboring vehicle can forward Figure 2. The taxonomy of vehicular ad hoc networks 209 Routing Protocols in Vehicular Ad Hoc Networks Figure 3. Unicast routing protocols packets. Packets should be delivered by carryand-forward scheme. The delay time is greatly growing if the multi-hop forwarding can not be utilized. Efforts will be made as finding a routing path with multi-hop forwarding. The min-delay routing protocols (Lochert et al., 2005; Zhao et al., 2006; Naumov et al., 2007; Chen et al., 2009; Taleb et al., 2007; Sun et al., 2006; Wan et al., 2008) are reviewed as follows. Greedy Perimeter Coordinator Routing Protocol Lochert et al. (2005) proposed GPCR (greedy perimeter coordinator routing) which is a positionbased routing for urban environment. The basic idea just likes Greedy Perimeter Stateless Routing (GPSR) (Karp et al., 2000). The routing path is established by the greedy method. GPCR protocol is very well suited for highly dynamic environments such as inter-vehicle communication on the highway or city. However, urban environment has the radio obstacles problem (Karp et al., 2001). Radio obstacles problem has the influence on the performance of position-based routing. Therefore, GPCR overcomes this problem by establishing robust routes in the city environment. GPCR adopts the similar idea of GPSR (Blaevi´c et al., 2002) for the selection of intermediate nodes. GPSR needs the global knowledge of the city topology with the street map. The sender determines which junctions have to be traversed by some control 210 packets. Forwarding between junctions is then done in a position-based fashion. Observe that, GPCR does not need the global knowledge of the street map. GPCR traverses the junctions by a restricted greedy forwarding procedure, and adjusts the routing path by the repair strategy which is based on the topology of streets and junctions. Figure 4 shows that vehicle Vu tries to send packets to vehicle VD. Vehicle 1a is selected as the next hop of Vu if greedy forwarding scheme is used. After vehicle 1a received the packets, vehicle 1a detects destination VD is not located at north. Vehicle 1a then moves packets backward vehicle 2a, then the packet is forwarded to VD. It shows that GPCR not require the global knowledge of the city map. Figure 4. Geographic routing protocol Routing Protocols in Vehicular Ad Hoc Networks VADD: Vehicle-Assisted Date Delivery Routing Protocol Data delivery routing protocol is developed by Zhao et al. (2006). They proposed a vehicleassisted data delivery protocol, called VADD, in VANETs. The VADD protocol adopted the idea of carry-and-forward for the data delivery. The most important issue is to select a forwarding path with the smallest packet delivery delay. It usually chooses the next hop closer to the destination for the data delivery, but this strategy is not suitable for the sparsely connected vehicular networks. In the sparsely connected vehicular networks, transmission loss commonly occurs in case of disconnection, the packet has to be carried by the vehicle, while the moving speed is significantly slower than multi-hop forwarding. To keep the low data transmission delay, VADD protocol transmits packets through wireless channels as much as possible, and if the packet has to be carried through roads, the road with higher speed is chosen firstly. VADD protocol assumes that vehicles are equipped with pre-loaded digital maps, which provide street-level map and traffic statistics such as traffic density and vehicle speed on roads at different times of the day. According to the information provided by digital maps, VADD protocol proposed a delay model to estimate the data delivery delay in different roads as follows, ( dij = 1 - e -R´rij )´ lij ´ c R +e -R´rij ´ lij uij , where dij is the expected packet forwarding delay from intersection Ii to intersection Ij, R is the communication range of vehicle, c is a constant used to adjust expected packet forwarding delay to a more reasonable value, rij, is the road from intersection Ii to intersection Ij, ρij is the vehicle density on rij, lij is the Euclidean distance of rij, and uij is the average vehicle velocity on rij. This delay model is very useful to estimate the data delivery delay in VANETs. With the delay model, VADD protocol estimates the best road with the lowest data delivery delay based on the current kept traffic patterns. In fact, it is difficult to find a road with the minimum forwarding delay between two arbitrary intersections from unlimited unknown intersections. In addition, observe that VADD only considers how to find a path from a mobile vehicle to a fixed location of destination vehicle. However, it is not easily collect the in time traffic pattern and information. From the out-of-date traffic information, VADD may establish the inappropriate road with the greater data delivery delay. Figure 5 illustrates that vehicle Va tries to send packets to the coffee shop, while the coffee shop is at the fixed location. Intersections Ia, Ib, Ic, and Id, are considered as the candidate intermediate intersections. After evaluating the expected forwarding delay, intersections Ia, Ic, and Id are chosen. This is because that the density of vehicle is high between intersections Ia, Ic, and Id, although it is not the shortest path. Connectivity-Aware Routing Protocol To overcome the limitation of the fixed location of destination vehicle, Naumov et al. (2007) proposed Connectivity-Aware Routing (CAR) protocol in VANETs. The CAR protocol is a position-based routing scheme. CAR protocol establishes a routing path from source to destination by setting the anchor points at intermediate intersections. Each vehicle exchanges hello messages to collect the information of neighboring vehicle, such as moving direction and speed. CAR sends the searching packets by using PGB algorithm (Preferred Group Broadcast) (Naumov et al., 2006) to find the destination and a routing path from source to destination. Each forwarding vehicle records it’s ID, hop count, and average number of neighbors 211 Routing Protocols in Vehicular Ad Hoc Networks Figure 5. The environment for VADD routing protocol in searching packets. Once the searching packets reach the destination, the destination chooses a better routing path which has the minimum delivery delay time and sends the reply packet to source. While destination sends the reply packet to source, the intersections passed through by the reply packet are set as the anchor point. After the path set up, data packets are forwarded in a greedy method toward the destination through the set of anchor points using the AGF algorithm (Advanced Greedy Forwarding) (Naumov et al., 2006). To maintain the routing path, a guard node is used if the destination changes the position. The guard node indicates the routing path to destination. Figure 6 gives that vehicle VS tries to send data to vehicle VD, the anchor points are set at I1,1, I2,1, I2,2, I3,2 and I3,4 . Data is forwarded according to order in the list of anchor points. DIR: Diagonal-IntersectionBased Routing Protocol To improve the CAR protocol, Chen et al. (2009) developed a diagonal-intersection-based routing 212 (DIR) protocol in VNETS. The key difference of CAR and DIR protocols is that DIR protocol (Chen et al., 2009) constructs a series of diagonal intersections between the source and destination vehicles. The DIR protocol is a geographic routing protocol. Based on the geographic routing protocol, source vehicle geographically forwards the data packet toward the first diagonal intersection, the second diagonal intersection, and so on, until the last diagonal intersection, and finally geographically reaches to the destination vehicle. For given a pair of neighboring diagonal intersections, two or more disjoint sub-paths exist between them. The novel property of DIR protocol is the auto-adjustability; while the auto-adjustability is achieved that one sub-path with low data packet delay, between two neighboring diagonal intersections, is dynamically selected to forward data packets. To reduce the data packet delay, the route is automatically re-routed by the selected sub-path with lowest delay. Figure 7 shows that DIR protocol constructs a series of diagonal intersections between vehicles VS and VD. Observe that, DIR protocol may set the fewer number of anchors Routing Protocols in Vehicular Ad Hoc Networks Figure 6. CAR is an anchor-based routing protocol than CAR protocol (Naumov et al., 2007). DIR protocol can automatically adjust routing path for keeping the lower packet delay, compared to CAR protocol (Naumov et al., 2007). ROMSGP Routing Protocol To improve the routing reliability, Taleb et al. (2007) proposed ROMSGP (Receive on Most Stable Group-Path) routing protocol in a city environment. Taleb et al. (2007) indicate that an unstable routing usually disconnects due to the loss of connectivity. This dis-connectivity is occurred if one vehicle moves out of the transmission range of a neighboring vehicle. A large difference of velocity exists between a pair of two vehicles. In ROMSGP protocol, all vehicles are split into four groups based on the velocity vector. A routing is said as a stable routing if the two vehicles are categorized in the same group. If two vehicles are categorized in different groups, the routing is an unstable routing. Each group has a unique ID and a base vector. A vehicle belongs to a group if the velocity vector has the maximum projection vector with this group. The stability of routing is evaluated by comparing the group ID if other vehicles receive packets from this sender vehicle. Figure 8 illustrates the ROMSGP routing protocol. Two routing paths are established, {VAVB, VBVD} and {VAVC, VCVD}. If VA, VB, VC, and VD belong to the same group, the two routing paths are both stable. Packet is delivered via {VAVB, VBVD} or {VAVC, VCVD}. If VB turns into another road, the projection vector is changed. VB belongs to the other group. Then the routing path {VAVC, VCVD} is the only choice. Reliable Routing for Roadside to Vehicle Communications In contrast with routing results developed in the highway or the city environments, it is very interest that Wan et al. (2008) specially proposed a reliable routing protocol in the rural environment. Wan et al. (2008) proposed two reliable routing strategies for roadside to vehicle (R2V) communication. The challenge of R2V communication in the rural environment is the terrain factor. For instance, a vehicle moving along the rural highway occasionally loses the line of sight (LOS) to the neighbor vehicle or to access points (APs) due to the obstacle-property caused by 213 Routing Protocols in Vehicular Ad Hoc Networks Figure 7. DIR protocol sets fewer anchors than CAR protocol Figure 8. ROMSGP routing protocol the curve roadway and mountains. In addition, almost no fixed communication infrastructure is available. Multi-hop inter-vehicle communication connecting to AP is the main solution of the R2V communication. High mobility causes the temporary network fragmentation problem. The link lifetime is very important issue for designing the reliable routing to overcome the temporary network fragmentation problem. The link lifetime is predicted by two conditions. Once the com- 214 munication is established, the link lifetime halts if (1) LOS between a pair of vehicles is lost, or (2) one vehicle moves out of the communication range of the neighboring vehicle. A link established in a shorter distance usually has longer link lifetime. The link lifetime is used to predict the lifetime of a route. A route is constructed by a series of links. The lifetime of a route is the minimum link lifetime in a route. Long lifetime of a route improves the routing reliability Routing Protocols in Vehicular Ad Hoc Networks if considered the lifetime-bounded shortest path. In addition to the lifetime of a routing path, the length-bounded maximum lifetime path is considered. To construct a length-bounded maximum lifetime path, reducing hops can improve the delivery delay-time. A routing path with fewer hops means the links are established in the long distance. Establishing a routing path with longer lifetime implies that the length of this routing path is long. Figure 9 (a) illustrates the example of lifetimebounded shortest path. The dotted line is current routing path and the link lifetime is going to end, where the minimum link lifetime is 9. The solid line is the candidate path. The link lifetime of solid line is greater than the threshold (=16). The routing path changes to solid line by AP assignment. Figure 9(b) illustrates the example of length-bounded maximum lifetime path. The dotted line is the routing path with minimum hops to AP (hops=4). The solid line is the selected path (hops=5). GVGRID: A QoS Routing Protocol To improve delivery delay-time and routing reliability, Sun et al. proposed GVGrid protocol (Sun et al., 2006) which is a QoS routing protocol for VANETs. GVGrid constructs a routing path from source to destination by grid-based approach. The goal of GVGrid is to maintain a high quality routing path between roadsides and vehicles, or between vehicles. GVGrid assumes each vehicle has digital map and location information. GVGrid divides the map into several grids. The RREQ and RREP packets are delivered through different grid to find a routing path through minimum number of grid. A grid is chosen based on the direction and the distance between vehicle and intersection and is selected as next grid if the direction of grid is the same as current grid or the grid is closed to the intersection. Then the intermediate grids between source and destination are recorded in the routing table. An appropriate vehicle which has the fewest number of disconnections in each grid is chosen to forward packets to next grid. A formula of evaluating the expected number of disconnections is derived in (Sun et al., 2006). The routing table records in terms of the source vehicle, destined grid, an appropriate vehicle as next hop with minimum the expected number of disconnections, a vehicle as previous hop, and the grid sequence. Once the routing path is broken, GVGrid just finds another vehicle in the grid instead of the previous vehicle. The routing path does not require finding again. Figure 10 (a) shows that vehicle VS floods RREQ message to find vehicle VD and vehicle VD replies RREP message to VS. Figure 10 (b) demonstrates that not only the grid sequence but also the information of the next vehicle are recorded in the routing table. Figure 9. (a) lifetime-bounded shortest path (b) length-bounded maximum lifetime path 215 Routing Protocols in Vehicular Ad Hoc Networks Figure 10.(a) An example of GVGrid routing protocol, (b) routing tables on vehicle VA and VD delay-bounded routing protocol Delay-bounded routing provides the low overhead in wireless communication. Delay-bounded routing delivers packet using carry-and-forward method as much as possible. Delay-bounded routing usually applies to non-real time services, such as FTP, e-mail, and HTTP services. The benefits of delay-bounded routing are follows; (1) the wireless media resource can be reserved for other real time service and emergency applications, and (2) packet delivery does not be affected by the network density. Packets can be delivered by carry-and-forward method or multi-hop forwarding depended on the constrained delay time. If the current delay time is less than the constrained delay time, carry-and-forward method is used. Otherwise, multi-hop forwarding method is used to quickly transmit packets. Skordylis et al. (2008) proposed a delaybounded routing protocol in VANETs, which provides a routing scheme that satisfy user-defined delay requirements while at the same time maintaining a low level of channel utilization. The delay-bounded routing protocol (Skordylis et al., 216 2008) focuses on the development of carry-andforward schemes that attempts to deliver data from vehicles to fixed infrastructure access point in an urban environment. Two routing algorithms, D-Greedy (Delay-bounded Greedy Forwarding) and D-MinCost (Delay-bounded Min-Cost Forwarding), evaluate traffic information and the bounded delay-time to carefully opt between the Data Muling and Multihop Forwarding strategies to minimize communication overhead while satisfying with the delay constraints imposed by the application. D-Greedy algorithm adopts only local traffic information to make routing decisions. D-Greedy algorithm chooses the shortest path to destined AP form the map information, and then allocates the constrained delay-time to each street within the shortest path according to the length of street. If packets can be delivered under the constrained delay-time in a street, Data Muling strategy is utilized. Packets are carried by a vehicle and forwarded at the vehicle’s speed to destined AP. Otherwise, Multihop Forwarding strategy is applied if packets cannot be delivered within the constrained delay-time. Packets are delivered Routing Protocols in Vehicular Ad Hoc Networks by multi-hop forwarding. To determine which strategy is adopted, the bounded delay time Del is evaluated as follows, defer_tim e(R ) = MaxDeferTime ´ (r - d ) SR r where TTL is the life time of packet, distToInt is the remaining length, and disToAP is the distance to AP station. The expected delay time at different time is evaluated by the equation below, DelDM = distToInt u A vehicle transmits packets to AP station by carry-and-forward method if Del>DelDM. If Del<DelDM, multi-hop forwarding is applied. D-MinCost algorithm considers the global traffic information in a city to achieve the minimum channel utilization within the constrained delay-time. According to the global traffic information, the cost C and delay Del of each street can be pre-computed. The cost C represents the number of message transmissions in a street. The delay Del denotes the time required to forward a message in a street. The cost C and delay Del of a street are evaluated as follows, DelDM = C MH = l ,C = 1, u DM 1 , DelMH + C MH ´ q , R where DelDM denotes the time required to carryand-forward a message through a street by Data Muling strategy, and l denotes the length of a street and u denotes the average vehicle velocity in this street. When the Data Muling strategy is applied, the packet is carried by a vehicle to forward, the vehicle only delivers the packet once at next intersection. The cost of Data Muling strategy is 1, so CDM = 1. The CMH represents the number of message transmissions in a street by Multihop Forwarding strategy. The R denotes the maximum transmission range of a vehicle. The q denotes the time required for a vehicle to check its neighbor list and identify the best next hop. To achieve the minimum cost C within the constrained delay Del, DSA (Delay Scaling Algorithm) (Goel et al., 2001) is applied, in order to efficiently compute delay-constrained least cost paths from the vehicle’s location to all access points on a network. The best routing path is opted by DSA (Goel et al., 2001) with minimum channel utilization under the constrained delaytime. Figure 11 shows that Data Muling strategy is applied if the packet can be delivered form VA to AP within the constrained delay-time. Otherwise, the packet is delivered by Multihop Forwarding strategy. When Multihop Forwarding strategy is applied, the packet is transmitted to a vehicle which is the closest to AP. Vehicle VC is choose by VA to relay packets to AP station. mulTicasT and geocasT rouTing proTocol Multicast and geocast routings are the other important routing operations in VANETs. Many VANET applications require disseminating information to a group of mobile vehicles in a VANET. These applications include distributed games, replicated file systems, teleconferencing, etc. The multicast in VANETs is defined by delivering multicast packets from a mobile vehicle to all multicast-member vehicles. The geocast in VANETs is defined by delivering geocast packets from a source vehicle to vehicles located in a specific geographic region. One of the challenges is how to develop the efficient multicast and geocast protocol over VANETs with the highly changeable topology and temporary network fragmentation to guarantee that 217 Routing Protocols in Vehicular Ad Hoc Networks Figure 11. Data muling and multihop forwarding strategies vehicles can be successfully received the multicast and geocast packets in VANETs. Usually, the temporary network fragmentation problem affects the performance of multicast and geocast operations. When a vehicle moves under a highly speed, the velocity variation between each pair of vehicles is large. The distance between each pair of vehicles quickly changes. This easily offers the temporary network fragmentation problem. When a vehicle easily moves out of the communication range of the sender and no neighboring vehicle is able to forward packets, then this vehicle fails to receive multicast and geocast packets. This condition is to lose the connectivity. Some results (Joshi et al., 2007; Bachir et al., 2003; Chen et al., 2009) have recently investigated the multicast and geocast protocols in a VANET. All of them mainly consider overcoming the temporary network fragmentation problem. It is very important to consider the factors of temporary network fragmentation problem when designing the multicast and geocast protocols. According to the property of geographic region, existing results can be classified into multicast, geocast protocol, and spatiotemporary multicast/geocast routing protocols as shown in Figure 12. This section reviews the existing results for VANETs. distributed robust geocast multicast routing protocol The goal of distributed robust geocast multicast routing protocol is to deliver packets to vehicles located in a specific geographic region. A vehicle should receive packets or drop only depended on its current location. If a vehicle is located in the specific geographic region, this vehicle receives packets. Otherwise this vehicle drops packets. Packet reception in the geocast multicast routing protocols mainly considers whether vehicles are located in the specific geographic region. Joshi et al. (2007) had proposed a distributed robust Figure 12.Existing multicast and geocast routing protocols 218 Routing Protocols in Vehicular Ad Hoc Networks geocast protocol for inter-vehicle communication. The zone of relevance (ZOR) is defined in (Joshi et al., 2007) as a geographic region which vehicles in this region should receive the geocast messages. Joshi et al. pointed out that a stable route is difficult to setup in a VANET due to limited lifetime of a connectivity and frequently network fragmentation. To enhance the reliability of receiving geocast messages under frequently changeable topology, the zone of forwarding (ZOF) is defined in (Joshi et al., 2007) as the geographic region which vehicles in this region should forward the geocast messages to other vehicles in the ZOR. Note that, ZOF usually surrounds ZOR to ensure the geocast messages can be delivered to vehicles inside ZOR. A periodic retransmission mechanism is proposed in (Joshi et al., 2007) to overcome the network fragmentation. Besides, a distance-based backoff algorithm is used to reduce the number of hops and redundant broadcasts. The random backoff time is determined by the distance to sender vehicle. The longer distance is, the smaller backoff time will be, the hops can be reduced. Once a vehicle broadcasts, the neighboring vehicles may not broadcast, redundant broadcasts can be reduced. Figure 13 shows that the temporary network fragmentation problem is overcame with the assistances of vehicles of ZOF, for instance, VG and VF. multicast protocol in ad hoc networks inter-Vehicle geocast Bachir et al. (2003) proposed a multicast protocol in ad hoc networks inter-vehicle geocast, called IVG protocol (Bachir et al., 2003). The IVG protocol is used to inform all the vehicles in a highway if any danger is occurred; such as an accident. The risk area is determined in terms of driving direction and positioning of vehicles. Vehicles located in the risk area form a multicast group. The multicast group is defined temporarily and dynamically by the location, speed, and driving direction of vehicles. IVG protocol uses periodic broadcasts to overcome temporary network fragmentation for delivering messages to multicast members. The re-broadcast period is calculated based on the maximum vehicle speed. Besides, IVG protocol reduces the hops of delivering message by using the deferring time, such that defer_tim e(R ) = MaxDeferTime ´ (r - d ) SR r Figure 13. DRG multicast routing protocol 219 Routing Protocols in Vehicular Ad Hoc Networks where r is the transmission range and dSR is the distance between sender vehicle VS and receiving vehicle VR. The farthest vehicle with larger dSR waits for the less time to re-broadcast. Figure 14 shows an example for the IVG protocol. Vehicle VA encounters car-function-failure problem and sends this notification to all vehicles in the risk area. Vehicles VB, VC, and VD form a multicast group since they are located in the risk area. Vehicle VC is the next hop of VA since the VC is farther from VA than VB. After vehicle VC sending out packets, vehicle VB not forwards packet. spatiotemporary multicast/ geocast routing protocol The spatiotemporary multicast/geocast routing protocol is a new and very interest routing problem. Unlike ordinary multicast and geocast routing protocols, the spatiotemporary multicast and geocast Figure 14. IVG multicast protocol Figure 15. Applications for mobicast routing 220 routing protocol should take the time factor into account. The distinctive feature of this new form of spatiotemporary multicast and geocast routing protocol is the delivery of information to all nodes that happen to be in a prescribed region of space at a particular point in time. Chen et al. (2009) presents a “spatiotemporary multicast”, called a “mobicast”, protocol for supporting applications which require spatiotemporary coordination in VANETs. The spatiotemporary character of a mobicast is to forward a mobicast message to vehicles located in some geographic zone at time t, where the geographic zone is denoted as zone of relevance (ZORt). Vehicles located in ZORt at the time t should receive the mobicast message. Many interesting and useful applications on VANETs can be supported by mobicast routing protocol, such as emergency event (Park et al., 2006), online game (Palazzi et al., 2007), and video advertisement (Yoon et al., 2008). Mobicast routing protocol can be effectively used for those VANETs applications for emergency warning, online game invitation, and video advertisement, as illustrated in Figure 15. Figure 15 (a) shows vehicle Ve has the control failure problem for a short period of time, and warning messages are sent to all nearby vehicles to avoid accident. By mobicast routing protocol, warning messages can be sent to vehicles in the warning area to avoid accident. Figure 15 (b) shows an online game application, vehicle Ve can invite those nearby vehicles to be game-playing members for a longer period of Routing Protocols in Vehicular Ad Hoc Networks time. Our mobicast routing protocol can make sure that the game information can disseminate to all members in ZORt. To ensure the mobicast message can be sent to all vehicles in ZORt, vehicles located in ZORt at the time t must keep the connectivity to maintain the real-time data communication between all vehicles in ZORt. The connectivity is kept of all vehicles in ZORt through the vehicular ad hoc networks. The connectivity of ZORt is lost if any vehicle in ZORt suddenly accelerates or decelerates its velocity. The temporal network fragmentation problem is occurred such that vehicle in ZORt cannot successfully receive the mobicast messages. To solve the problem, Chen et al. proposed a new mobicast protocol to successfully disseminate mobicast messages to all vehicles in ZORt via a special geographic zone, called as zone of forwarding (ZORt). The proposed mobicast routing protocol can dynamically estimate the accurate ZORt. to successfully disseminate mobicast messages to all vehicles in ZORt. As shown in Figure 16, gives a continuous-time example of ZORi, where i = t…t + 2, with the temporal network fragmentation problem. The transmission range of each vehicle is assumed to r. Initially, Ve detects an emergency event at time t to form a ZORt. V1 and V3 are involved by this event and indicated to receive the mobicast message by ZORt. V1 and V3 are located within the transmission range of Ve; thus Ve directly disseminates the mobicast message to V1 and V3. At time t + 1, V2 and V4 approach Ve and they are indicated to receive the mobicast message by ZORt+1. Observed that although V4. is out of transmission range of Ve, V4 can receive the mobicast message by relaying from V1. At time t + 2, V1 moves away from V4 and V2 moves out of transmission range of Ve; thus V2 and V4 can not receive the mobicast message. The temporal network fragmentation problem occurred on V2 and V4. ZORt is introduced below to solve this problem. To overcome the temporal network fragmentation problem, ZORt is used to disseminate the mobicast message to all vehicles located in the ZORt. ZORt is a geographic zone greater than ZORt to involve more vehicles to forward the mobicast message. Vehicles in ZORt should forward the mobicast message to another vehicles located in ZORt. ZORt indicates which vehicle should forward the mobicast message to other vehicles located in the ZORt. All vehicles in ZORt must forward the received mobicast message; even those vehicles are not located in ZORt. Figure 17(a) shows V2 and V4 can not receive the mobicast message due to the temporal network fragmentation problem. Example of ZORt is illustrated in Figure 17(b), V5 and V6 are located in ZORt and have the responsibility of forwarding the mobicast message to V2 and V4, respectively. Normally, the size of ZORt may be larger or smaller than the optimal size of ZORt . If the size of ZORt is larger than the optimal size of ZORt, some irrelevant vehicles are asked to uselessly forward the mobicast message. If the size Figure 16. Temporal network fragmentation problem 221 Routing Protocols in Vehicular Ad Hoc Networks of ZORt is smaller than the optimal size of ZORt, the temporal network fragmentation problem is incompletely overcome. Observe that, the size of ZORt is difficult to predict and determined under the high speed environment, such that it easily wastes the network resources. Chen et al. proposed an efficient scheme to estimate the size of ZORt is near to the optimal size of ZORt. Therefore, zone V V of approaching ZOAt i or Zt i is proposed herein to accurately predict the ZORt. V V ZOAt i or Zt i is an elliptic zone of approaching to forward the mobicast message more closed V to a destined vehicle and Zt i is initiated by vehicle V Vi at time t. Any vehicle in the Zt i has the responsibility of forwarding the mobicast message sent V from vehicle Ve. Zt i bounds the mobicast message V propagation, vehicles in the Zt i can only forward the mobicast message to other vehicles located in V the Zt i . If a vehicle cannot successfully forward the mobicast message to any neighbor vehicle V in the Zt i which is more closed to the destined vehicle, a new approaching zone is initiated. The Vi from growing of a new zone of approaching Zt +1 V an existing zone of approaching Zt i is explained as follows. Given two connected approaching zones V V i Zt i and ZtV+1 , if Vi+1 in Zt i cannot successfully forward the mobicast message to any neighbor vehicle closed to the destined vehicle, then Vi+1 Vi , where initiates a new zone of approaching Zt +1 Figure 17. Function of ZORt and ZOFt 222 V i . Therefore, multiple zones Vi+1 is located in Zt +1 of approaching are initiated to forward the mobicast message, such that ZORt is finally formed by all initiated zones of approaching. Therefore, we Vi V V ∪… ∪ have ZORt = ZORt∪ Zt i ∪… ∪ Zt i ∪ Zt +1 Vn Zt . Observe that, ZORt is the partial ZORt since the mobicast message should be transmitted to all vehicles located in ZORt. Figure 18 demonstrates an example of the proposed mobicast routing protocol. At time t, V1, V2, V3, and V4 are located in ZORt and receive the mobicast message from vehicle Ve. At time t +1, V2 and V4 can not directly receive the mobicast messages due to temporal network fragmentation problem. At time t +1, Ve, Ve V5 V1 , Zt +1 , and Zt +1 to forward V5, and V1 initiate Zt +1 the mobicast messages to vehicles V4 and V2. In Ve V5 V1 ∪ Zt +1 ∪ Zt +1 . this case, ZORt+1 = ZORt+1∪ Zt +1 broadcasT rouTing proTocol Broadcast is the last important operation for a vehicle to disseminate a broadcast message to all the others in a VANET. To overcome the temporary network fragmentation problem to avoid some vehicles cannot be successfully received the broadcast message. This section describes existing broadcast routing protocols in VANETs, as listed in Figure 19, as follows. Routing Protocols in Vehicular Ad Hoc Networks Figure 18. Zone of approaching ZORt Figure 19. Existing broadcast routing protocols on The broadcast storm problem in ad hoc wireless networks The broadcast storm problem has been heavily discussed in MANETs (Tseng et al., 2002). Tonguz et al. (2006) further demonstrates the broadcast storm problem occurred in VANETs. The broadcast storm problem causes serious packet collision and packet loss since too many vehicles simultaneously broadcast messages. A shortest routing path can be discovered by broadcasting the RREQ (route request) message. The broadcast routing strategies developed in (Tonguz et al., 2006) are 1-persistence and p-persistence. However, these two strategies easily cause the broadcast storm problem, especially in a high density network. Tonguz et al. (2006) thus proposed three distributed broadcast suppression techniques, weighted p-persistence, slotted 1-persistence, and slotted p- persistence schemes. In the weighted p-persistence scheme, if vehicle Vj receives a packet form vehicle Vi, vehicle Vj first checks whether the packet has been received. If vehicle Vj receives this packet at the first time, then vehicle Vj has probability pij to re-broadcast the packet. Otherwise, vehicle Dij , where Dij Vj drops this packet, under pij = R is the distance between vehicle Vi and Vj, R is the transmission range. Neighbors of vehicle Vi change pij to 1 to ensure that the message must be broadcasted if they have no received the rebroadcast message after waiting a random time. In the slotted 1-persistence scheme, If vehicle Vj firstly receives this packet from vehicle Vi, then vehicle Vj waits for TS time slots, vehicle Vj has ij probability 1 to re-broadcast the packet, where TS = Sij×τ. ij 223 Routing Protocols in Vehicular Ad Hoc Networks ïìïé æ D öù ïïêêN ççç1 - ij ÷÷÷úú , Sij = íê s çè R ÷÷øú ú ïïê ïï 0, ïî if dV-casT: broadcasting in VaneT Dij if , Dij > R where τ is the propagation time for one hop transmission and Ns is the default number of time-slot. The slotted p-persistence scheme combines the weighted p-persistence and slotted 1-persistence schemes. If vehicle Vj firstly receives the packet from Vi, then vehicle Vj waits for TS time-slots. ij Vehicle Vj has probability pij to re-broadcast the packet. Simulation results show that these three schemes can achieve up to 90% reduction of packet loss rate. Figure 20 gives the example of these three broadcast schemes. In Figure 20(a), the weighted p-persistence scheme is used. Figure 20(b) illustrates the slotted 1-persistence scheme. Figure 20(c) shows the slotted p-persistence scheme. Tonguz et al. (2007) proposed DV-CAST for a multi-hop broadcast routing protocol in VANETs. Tonguz et al. (2007) indicates three traffic scenarios for a vehicular broadcasting; (1) dense traffic scenario, (2) sparse traffic scenario, and (3) regular traffic scenario. Tonguz et al. (2007) integrates previously proposed routing solution in (Tonguz et al., 2006) to develop DV-CAST which is suitable for both of dense and sparse traffic scenarios. The overhead in dense traffic scenario can be reduced. DV-CAST utilizes the information form one-hop neighbors to make the routing decision. Each vehicle records the states of neighboring vehicles all the time. Three parameters should be recorded; (1) DFlag (Destination Flag), (2) MDC (Message Direction Connectivity), and (3) ODC (Opposite Direction Connectivity). The DFlag indicates whether the direction of the current vehicle is the same as the source vehicle. The MDC describes whether a vehicle exists behind Figure 20. Broadcast suppression techniques (a) weighted p-persistence scheme (b) slotted 1-persistence scheme (c) slotted p-persistence scheme 224 Routing Protocols in Vehicular Ad Hoc Networks Figure 21. Well-connected neighborhood the current vehicle to assist broadcast message forwarding. The ODC indicates whether a vehicle exists in the opposite direction. Vehicles can make decision based on these three parameters. For example, if a vehicle Vi receives a new broadcast message, vehicle Vi firstly checks MDC. If there are vehicles behind vehicle Vi(MDC=1), vehicle Vi uses broadcast suppression schemes proposed in (Tonguz et al., 2006) to forwarding the broadcast message. Otherwise (MDC=0), vehicle Vi checks ODC. If there are vehicles in the opposite direction (ODC=1, MDC=0), vehicle Vi forwards the broadcast message to vehicles in the opposite direction. After vehicle Vi broadcasting message, vehicle Vi checks DFlag. If the direction of vehicle Vi is different from source vehicle (DFlag=0), vehicle Vi overhears for a period of time to ensure that the message is successfully broadcasted. Figure 21 shows that the broadcast message is initiated by VS and it is forwarded from group 1 to group 2. Although groups 1, 2, and 3 are dense group, groups 1 and 2 encounter the temporary network fragmentation problem. Group 1 cannot directly forward packets to group 2. In this case, vehicle VS can forward packets to group 3 which is in the reverse direction, then vehicle VS forwards packets to group 2. Observe that, the temporary network fragmentation problem is also considered in the design of broadcasting. broadcast methods for inter-Vehicle communications system Fukuhara et al. (2005) proposed broadcast methods for inter-vehicle communications system to provide emergency information dissemination in VANETs. The purpose of emergency information is to announce an urgent event by broadcasting for surrounding vehicles. According to the purposes of emergency information, the proposed broadcast methods in (Fukuhara et al., 2005) are divided into two categories, emergency-vehicle-approach information and traffic accident information. Emergency-vehicle-approach information is used to announce the urgent event to those vehicles in front of the current vehicle, so the emergency information is only disseminated ahead. Traffic accident information is used to announce the urgent event to those vehicles behind the current vehicle; the emergency information is only disseminated behind. By limiting the broadcast direction, the proposed broadcast methods (Fukuhara et al., 2005) can provide broadcasts to a particular area and avoid mistakenly notifying other areas where the information is not needed. To decide emergency information should be received, forwarded, or dropped, three conditions below are examined: (X C - X E + aV F ×V F > 0 , ) X C - X E + aV F < Rnotification , (1) (2) 225 Routing Protocols in Vehicular Ad Hoc Networks Figure 22. The broadcast area for emergency information X C - X E + aV F < Rrelay , (3) where XC is the location vector of the current vehicle, XE is the location vector of the emergency vehicle, VF is the forward direction vector of the emergency information, Rnotification is the available notification range, and Rrelay is the available relay range. If conditions (1) and (2) are true, the current vehicle receives the emergency information. If conditions (1) and (3) are true, the current vehicle re-broadcasts the emergency information. Figure 22 shows that vehicle VA broadcasts the emergency message to the restricted direction. Vehicle VD does nothing. Vehicle VD is located in the relay range, it re-broadcasts the emergency information. Vehicle VD is located in notification range but not in relay range, VC just receives the emergency information and not to re-broadcast. conclusion Unicast, multicast, and broadcast routing operations are key issues in the network layer for VANETs. This chapter surveys existing unicast, multicast, and broadcast protocols for VANETs. The unicast routing protocols are split into min-delay and delay-bound approaches. The min-delay unicast routing protocols construct a minimum-delay routing path as soon as pos- 226 sible. The delay-bound routing protocol utilizes the carry-and-forward technique to minimize the channel utilization within a constrained delay time. This chapter also surveys important multicast and geocast protocols for VANETs. The multicast in VANETs is defined by delivering multicast packets from a mobile vehicle to all multicast-member vehicles. The geocast in VANETs is defined by delivering geocast packets from a source vehicle to vehicles located in a specific geographic region. A mobicast routing protocol in VANETs is also described. Finally, broadcast protocols in VANETs are also introduced. We predict the tendency of the design of routing protocols for VANETs must be the low communication overhead, the low time cost, and high adjustability for the city, highway, and rural environments. reFerences Amouris, K. (2005). Position-Based Broadcast TDMA Scheduling for Mobile Ad-Hoc Networks (MANETs) with Advantaged Nodes. IEEE Military Communications Conference, 1, 252–257. ASTM International E2213-03 (2003). Standard Specification for Telecommunications and Information Exchange Between Roadside and Vehicle Systems 5 GHz Band Dedicated Short Range Communications (DSRC) Medium Access Control (MAC) and Physical Layer (PHY) Specifications. Routing Protocols in Vehicular Ad Hoc Networks Bachir, A., & Benslimane, A. (2003). A Multicast Protocol in Ad hoc Networks Inter-Vehicle Geocast. IEEE Semiannual Vehicular Technology Conference, 4, 2456-2460. Blaževi’c, L., Giordano, S., & LeBoudec, J. Y. (2002). Self Organized Terminode Routing. Cluster Computing Journal, 5(2), 205–208. doi:10.1023/A:1013998030317 Blum, J., Eskandarian, A., & Hoffman, L. (2004). Challenges of inter-vehicle ad hoc networks. IEEE Transactions on Intelligent Transportation Systems, 5(4), 347–351. doi:10.1109/ TITS.2004.838218 Hughes, L., & Maghsoudlou, A. (2006). An Efficient Coverage-Based Flooding Scheme for Geocasting in Mobile Ad Hoc Networks. International Conference on Advanced Information Networking and Applications, 1, 6. Joshi, H. P., Sichitiu, M., & Kihl, M. (2007). Distributed Robust Geocast Multicast Routing for Inter-Vehicle Communication. WEIRD Workshop on WiMax, Wireless and Mobility. Karp, B. (2001). Challenges in Geographic Routing: Sparse Networks, Obstacles, and Traffic Provisioning. In DIMACS Workshop on Pervasive Networking. Briesemeister, L., & Hommel, G. (2000). Overcoming Fragmentation in Mobile Ad hoc Networks. Journal of Communications and Networks, 2(3), 182–187. Karp, B., & Kung, H. T. (2000). GPSR: Greedy Perimeter Stateless Routing for Wireless Networks. ACM International Conference on Mobile Computing and Networking, (pp. 243-254). Chen, Y. S., Lin, Y. W., & Lee, S. L. (2009). A Mobicast Routing Protocol for Vehicular Ad Hoc Networks. Mobile Networks and Applications (MONET).New York: ACM/Springer. Lochert, C., Mauve, M., Füßler, H., & Hartenstein, H. (2005). Geographic routing in city scenarios. ACM SIGMOBILE Mobile Computing and Communications, 9(1), 69–72. doi:10.1145/1055959.1055970 Chen, Y. S., Lin, Y. W., & Pan, C. Y. (2009). A Diagonal-Intersection-Based Routing Protocol for Urban Vehicular Ad Hoc Networks. Telecommunication System, (under revision). Fukuhara, T., Warabino, T., Ohseki, T., Saito, K., Sugiyama, K., Nishida, T., & Eguchi, K. (2005). Broadcast Methods for Inter-vehicle Communications System. IEEE Wireless Communications and Networking Conference, 4, 2252-2257. Gerten, G. (2005). Protecting the Global Positioning System. IEEE Aerospace and Electronic Systems Magazine, 20(11), 3–8. doi:10.1109/ MAES.2005.1576067 Goel, A., Ramakrishnan, K. G., Kataria, D., & Logothetis, D. (2001). Efficient Computation of Delay-Sensitive Routes from One Source to All Destinations. IEEE Conference on Computer Communications (INFOCOM), (pp. 854–858). Manoharan, R., & Thambidurai, S. L. P. P. (2008). Energy Efficient Robust On-Demand Multicast Routing Protocol for MANETs. [IJAHUC]. International Journal of Ad Hoc and Ubiquitous Computing, 3(2), 90–98. doi:10.1504/ IJAHUC.2008.017002 Naumov, V., Baumann, R., & Gross, T. (2006). An Evaluation Of Inter-Vehicle Ad Hoc Networks Based On Realistic Vehicular Traces. In ACM International Symposium on Mobile Ad Hoc Networking and Computing, (pp. 108-119). Naumov, V., & Gross, T. (2007). ConnectivityAware Routing (CAR) in Vehicular Ad Hoc Networks. IEEE International Conference on Computer Communications (INFOCOM’07), (pp. 1919-1927). 227 Routing Protocols in Vehicular Ad Hoc Networks Palazzi, C. E., Roccetti, M., Pau, G., & Gerla, M. (2007). Online Games on Wheels: Fast Game Event Delivery in Vehicular Ad-hoc Networks. In IEEE Intelligent Transportation Systems Society, (pp. 42–49). Park, J. S., Lee, U., Oh, S. Y., Gerla, M., & Lun, D. (2006). Emergency Related Video Streaming in VANETs using Network Coding. In ACM International Workshop on Vehicular Ad Hoc Networks (VANET), (pp. 102-103). Safa, H., Artail, H., & Shibli, R. (2009). An Interoperability Model for Supporting Reliability and Power-Efficient Routing in MANETs. [IJAHUC]. International Journal of Ad Hoc and Ubiquitous Computing, 4(2), 71–83. doi:10.1504/ IJAHUC.2009.023898 Saha, A. K., & Johnson, D. B. (2004). Modeling mobility for vehicular ad hoc networks. ACM International Workshop on Vehicular Ad Hoc Networks (VANET), (pp. 91-92). Sawamura, T., Tanaka, K., Atajanov, M., Matsumoto, N., & Yoshida, N. (2008). Adaptive Router Promotion and Group Forming in Ad-Hoc Networks. [IJAHUC]. International Journal of Ad Hoc and Ubiquitous Computing, 3(4), 217–223. doi:10.1504/IJAHUC.2008.018907 Skordylis, A., & Trigoni, N. (2008). DelayBounded Routing in Vehicular Ad-Hoc Networks. ACM international symposium on Mobile ad hoc networking and computing, (pp. 341-350). Sun, W., Yamaguchi, H., Yukimasa, K., & Kusumoto, S. (2006). GVGrid: A QoS Routing Protocol for Vehicular Ad Hoc Networks. IEEE International Workshop on Quality of Service, (pp. 130-139). Taleb, T., Sakhaee, E., Jamalipour, A., Hashimoto, K., Kato, N., & Nemoto, Y. (2007). A Stable Routing Protocol to Support ITS Services in VANET Networks. IEEE Transactions on Vehicular Technology, 56(6), 3337–3347. doi:10.1109/ TVT.2007.906873 228 Tonguz, O., Wisitpongphan, N., Bai, F., Mudalige, P., & Sadekar, V. (2007). Broadcasting in VANET. Mobile Networking for Vehicular Environments, (pp. 7-12). Tonguz, O. K., Wisitpongphan, N., Parikh, J. S., Bai, F., Mudalige, P., & Sadekar, V. K. (2006). On the Broadcast Storm Problem in Ad hoc Wireless Networks. International Conference on Broadband Communications, Networks and Systems, (pp. 1-11). Tseng, Y. C., Ni, S. Y., Chen, Y. S., & Sheu, J. P. (2002). The Broadcast Storm Problem in a Mobile Ad Hoc Network. [WINET]. ACM Wireless Networks, 8(2), 153–167. doi:10.1023/A:1013763825347 Wan, S., Tang, J., & Wolff, R. S. (2008). Reliable Routing for Roadside to Vehicle Communications in Rural Areas. In IEEE International Conference on Communications, (pp. 3017-3021). Xu, Q., Mak, T., Ko, J., & Sengupta, R. (2004). Vehicle-to-Vehicle Safety Messaging in DSRC. In ACM International Workshop on Vehicular Ad Hoc Networks (VANET), (pp. 19-28). Yoon, H., Kim, J., Tan, F., & Hsieh, R. (2008). On-demand Video Streaming in Mobile Opportunistic Networks. In IEEE International Conference on Pervasive Computing and Communications (PerCom), (pp. 80–89). Yousefi, S., Mousavi, M. S., & Fathy, M. (2006). Vehicular Ad Hoc Networks (VANETs): Challenges and Perspectives. International conference on ITS Telecommunications, (pp. 761-766). Zhao, J., & Cao, G. (2006). VADD: VehicleAssisted Data Delivery in Vehicular Ad Hoc Networks. IEEE Computer Communications, (pp. 1-12). 229 Chapter 14 Applications in Vehicular Ad Hoc Networks Tzung-Shi Chen National University of Tainan, Tainan, Taiwan, R.O.C. Hua-Wen Tsai National University of Tainan, Tainan, Taiwan, R.O.C. Yi-Shiang Chang National University of Tainan, Tainan, Taiwan, R.O.C. absTracT The various sensors and wireless communication devices have been extensively applied to daily life due to the advancements of microelectronics mechanism and wireless technologies. Recently, vehicular communication systems and applications become more and more important to people in daily life. Vehicular communication systems that can transmit and receive information to and from individual vehicles have the potential to significantly increase the safety of vehicular transportation, improve traffic flow on congested roads, and decrease the number of people of deaths and injuries in vehicular collisions effectively. This system relies on direct communication between vehicles to satisfy the communication needs of a large class of applications, such as collision avoidance, passing assistance, platooning. In addition, vehicular communication systems can be supplemented by roadside infrastructure to access Internet and other applications. This system forms a special case of mobile ad hoc networks called Vehicle Ad Hoc Networks (VANETs). They can be formed between vehicles with vehicle to vehicle (V2V) communication or between vehicles and an infrastructure with vehicle to infrastructure (V2I) communication. The applications and characteristics of VANETs are introduced and presented in this Chapter. inTroducTion Recently, the various sensors and wireless communication devices have been extensively applied to daily life due to the advancements of microDOI: 10.4018/978-1-60566-840-6.ch014 electronics mechanism and wireless technologies. Vehicular transportation is one of the main modes of transportation for millions of U.S. citizens and hundreds of millions around the world in spite of increasing oil prices and environmental concerns. However, 38,252 Americans were killed in 2003 with 2,697,000 seriously injured. Furthermore, Copyright © 2010, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global is prohibited. Applications in Vehicular Ad Hoc Networks each day the average American spends 2.5 hours in his/her vehicle, a significant percentage of this time in traffic jams and at stop lights. The statistics are similar in many other parts of the world (Sichitiu et al., 2008). On an average day in the United States, vehicular collisions kill 116 and injure 7900. More health care dollars are consumed in the United States treating crash victims than any other cause of illness or injury; the situation in the European Union is similar, with over 100 deaths and 4600 injuries daily, and the annual cost of € 160 billion. Governments and automotive companies are responding by making the reduction of vehicular fatalities a top priority (Robinson et al., 2006). Vehicular communication systems that can transmit and receive information to and from individual vehicles have the potential to significantly increase the safety of vehicular transportation, improve traffic flow on congested roads, and decrease the number of people of deaths and injuries in vehicular collisions effectively. This system relies on direct communication between vehicles to satisfy the communication needs of a large class of applications (e.g., collision avoidance, passing assistance, platooning, and so on). Vehicular communication systems can be supplemented by roadside infrastructure to access Internet and other applications. This system forms a special case of mobile ad hoc networks called Vehicle Ad Hoc Networks (VANETs). They can be formed between vehicles with vehicle to vehicle (V2V) communication or between vehicles and an infrastructure with vehicle to infrastructure (V2I) communication. VANETs can increase security, efficiency and comfortable trip in collecting road traffic. Such systems can enable a wide range of applications, such as collision avoidance, emergency message dissemination, dynamic route scheduling, route navigation, and real-time traffic condition monitoring. Traditional vehicular networks for reporting accidents or traffic conditions rely on certain infrastructure, such as road-side traffic sensors 230 reporting data to a central database, or cellular wireless communication between vehicles and a monitoring center. The advantage of VANET is not needed to deploy too many expensive infrastructures to be installed on every road and intersections. VANETs are emerging as the preferred network design for intelligent transportation systems. VANETs are based on short-range wireless communication (e.g., IEEE 802.11) between vehicles. However, the disadvantage of VANETs is in supporting complex network protocols. The existing researches focus on medium access control (MAC) and routing issues, in particular pointing to the mismatch between the need of inter-vehicle communication applications for group communications and the services offered by mobile ad hoc network (MANET) routing protocols. Several applications enabled by vehicular communication systems are classified. The most commonly considered applications are related to public safety and traffic coordination. Traffic management applications, traveler information support, and various comfort applications have the potential to make travelling more efficient, convenient and pleasant. For each type of application, we consider its addressing and real-time requirements and the type of vehicular communication necessary for its implementation. Generally speaking, vehicular communication systems are classified into three major types, inter-vehicle communication (IVC), roadsideto-vehicle communication systems (RVC), and hybrid vehicular communication (HVC). IVC systems are completely infrastructure-free and only need some in-vehicle equipments. Depending on whether the information is retransmitted at intermediate hops or not, we can further distinguish between single-hop and multi-hop IVCs, SIVCs and MIVCs, respectively. SIVC systems are useful for applications requiring short-range communications, e.g., lane merging, automatic cruise control. MIVC systems are more complex than SIVCs but can also support applications that require long-range communications, e.g., traffic Applications in Vehicular Ad Hoc Networks monitoring. An MIVC system requires a network layer capable of multi-hop routing. In RVC systems, all communications take place between roadside infrastructure and vehicles. Depending on the application, two different types of infrastructure can be classified, sparse RVC (SRVC) and ubiquitous RVC (URVC) systems. SRVC systems are capable of supporting communication services at hot spots, for examples, a busy intersection scheduling its traffic light, a restaurant advertising its existence and prices, and parking availability at an airport. An SRVC system can be deployed gradually, thus not requiring substantial investments before any available benefits. A URVC system is the holy grail of vehicular communication in which providing all roads with high-speed communication would enable applications unavailable with any of the other systems. Unfortunately, a URVC system may require considerable investments for providing full coverage of existing roadways, especially in large countries (Sichitiu et al., 2008). HVC systems are proposed for extending the range of RVC systems. In HVC systems vehicles communicate with roadside infrastructure even when they are not in direct wireless range by using other vehicles as mobile routers. The main advantage is that it requires less roadside infrastructure. However, one disadvantage is that network connectivity may not be guaranteed in scenarios with low vehicle density. The Federal Communications Commission (FCC) has allocated spectrum for IVC and similar applications (e.g., Wireless Access in Vehicle Environment, WAVE). Governments and prominent industrial corporations, such as Toyota, BMW, and Daimler-Chrysler have launched important projects for IVC communications. Some notable projects, e.g., Advanced Driver Assistance Systems (ADASE2), Crash Avoidance Metrics Partnership (CAMP), Chauffeur in EU, CarTALK2000, FleetNet, California Partners for Advanced Transit and Highways (California PATH), and DEMO 2000 by Japan Automobile Research Institute (JSK) are a major step towards the realization of Intelligent Transport Services (ITS) (Taleb et al., 2007). Dedicated short range communications (DSRC) standard at 5.9 GHz band is projected to support low-latency wireless data communications between vehicles and from vehicles to roadside units. The DSRC specification is meant to be an extension of the IEEE 802.11 technology into the outdoor high-speed vehicle environment. In fact, the physical layer (PHY) of DSRC is adapted from IEEE 802.11a PHY based on orthogonal frequency division multiplex (OFDM) technology. Moreover, the multiple access control (MAC) layer of DSRC is very similar to the IEEE 802.11 MAC based on the carrier sense multiple access with collision avoidance (CSMA/CA) protocol with some minor modifications (Ryu et al., 2004). MANETs are wireless multi-hop networks, decentralized and self-organized without infrastructure. VANETs satisfy all these requirements, and are a special class of MANETs. However, the followings are several characteristics that differentiate between VANETs and MANETs (Sichitiu et al., 2008) . • Applications: The most common assumption of MANET applications are identical to those enabled by the Internet. In contrast, VANETs have completely different applications to increase the safety and comfortable trip. VANET applications include onboard active safety systems to assist drivers in avoiding collisions and to coordinate among them at critical points such as intersections and highway entries. Safety systems may intelligently disseminate road information, such as incidents, real-time traffic congestion, high-speed tolling, or surface condition to vehicles in the vicinity of the subjected sites. This helps to avoid platoon vehicles and to accordingly improve the roads capacity. With such active safety systems, the number of 231 Applications in Vehicular Ad Hoc Networks • • • • car accidents and associated damage are expected to be largely reduced. Receiver: MANET applications require point-to-point (unicast) with fixed addressing. The recipient of a message is another node in the network specified by its IP address. However, VANET applications often require dissemination of the messages to many nodes (multicast) that satisfy some geographical constraints and possibly other criteria (e.g., direction of movement). The need for this addressing mode requires a significantly different routing paradigm. Rate of link changes: The nodes in MANETs are assumed to have moderate mobility. This assumption allows MANET routing protocols to establish end-to-end paths that are valid for a reasonable amount of time and only occasionally need repairs. In VANETs applications, it is shown that due to the high degree of mobility of the nodes involved, even multi-hop paths that only use nodes moving in the same direction on a highway have a lifetime comparable to the time needed to discover the path. Therefore, the communication of VANETs has more challenge than that of MANETs. Mobility model: In MANETs, the random waypoint (RWP) is the most commonly employed mobility model. However, for VANETs, most existing literatures recognized that RWP would be a very poor approximation of real vehicular mobility; instead, detailed vehicular traffic simulators are used. Energy efficiency: While in MANETs a significant body of literatures is concerned with power-efficient protocols, VANETs enjoys a practically unlimited power supply. In VANETs, the hardware available in each vehicle is listed below. 232 • • • • • A central processing unit (CPU) that implements the applications and communication protocols; A wireless transceiver that transmits and receives data to/from the neighboring vehicles and roadside units; A GPS receiver that provides relatively accurate positioning and time synchronization information; Appropriate sensors to measure the various parameters that have to be measured and eventually transmitted; An input/output interface that allows human interaction with the system. applicaTions in VaneTs Based on the different types of vehicular communication systems (e.g., IVC and RVC), the telematics have five different classifications in the applications shown in Figure 1 (Sichitiu et al., 2008). The applications in VANETs consist of the public safety application, traffic management applications, traffic coordination and traffic assistance, traveler information support, and comfort applications. The classifications of these applications are not meant to the overall applications in VANETs. However, the adopted classification can represent the greater part of the current general applications in VANETs. The explanations about the applications will be described as follows in detail. Public safety applications are very important issue in the Telematics. Safety applications are geared primarily toward avoiding accidents and loss of life of the occupants of vehicles. Collision warning systems have been proposed to reduce the number of vehicle collisions in several scenarios. On highways, frontal collisions with slow moving (or stopped) vehicles are one of the most common types of accidents, often with serious consequences. A vehicle with its airbags deployed, Applications in Vehicular Ad Hoc Networks Figure 1. Telematics applications or a stopped or rapidly decelerating vehicle can transmit warning signals to all other approaching vehicles (shown in Figure 2(a)). Intermediate relays may be used to increase the dissemination range of the warning beyond the direct transmission range. At junctions, vehicles running red stop lights often result in side crashes. If both vehicles to be involved in the accident are equipped with vehicular communication systems, such accidents can be prevented (shown in Figure 2(b)). A similar situation may occur with other types of vehicles (e.g., trains). In some cases, if a collision is imminent, the system may be able to prepare the vehicles for collision (inflate air bags, tighten seat belts, etc.). Thus, safety applications have obvious real-time constraints, as drivers have to be notified before the information is no longer useful. In terms of addressing, the destinations in these applications will not be individual vehicles, but rather any relevant vehicle. The zone of relevance (ZOR) (also known as the target area) is determined by the particular application. Figure 2. Safety applications 233 Applications in Vehicular Ad Hoc Networks For example, an accident in the right lane of a highway will only affect vehicles approaching the accident from behind, while at an intersection the vehicles with intersecting trajectories and speeds over a threshold are relevant. Traffic management applications are focused on improving traffic flow, thus reducing both congestion as well as accidents resulting from congestion, and reducing travel time. The traffic management applications include the traffic monitoring, traffic light scheduling, and emergency vehicles. Traffic monitoring can provide high-resolution localized timely information regarding the traffic for several miles around the current location of the vehicle. For this application each vehicle in the system will act as a sensor (determining its current speed), as a relay (if the information is to travel for more than the direct transmission range) as well as a destination (using information from the other vehicles in the system (shown in Figure 3(a))). The information can be used to simply inform the driver or, in more complex systems, to reroute, estimate the time to destination, or even control the traffic by using adaptive speed limits, ramp metering, and so on. Emergency vehicles may notify the relevant vehicles as well as equipped stop lights of their Figure 3. Traffic management applications 234 presence and intended routes (shown in Figure 3(b)). Traffic light scheduling can be significantly improved by using an SRVC (Sparse Roadsideto-Vehicle Communication) system. Currently, many traffic lights are scheduled either statically or only considering limited information (e.g., by sensing the presence or absence of a vehicle in front of a traffic light). An SRVC system can provide additional information, such as the length of the queues at the traffic light as well as the number of vehicles expected to arrive in the near future, which can improve the efficiency of schedules. Applications in this class generally do not have stringent real-time requirements: the quality of the information degrades gracefully with the increase of delay and packet loss. Similar to the case of safety applications, the destinations of the information are any vehicles in the ZOR. For traffic monitoring applications the ZOR can be several miles around the information source. For traffic light scheduling, traffic lights being approached are appropriate destinations. Traffic coordination and traffic assistance have been the main research topics of many IVC projects. Platooning (i.e., forming tight columns of vehicles closely following each other on highways) has the potential to radically increase the capacity of existing highways. High-speed closed Applications in Vehicular Ad Hoc Networks loop control is of paramount importance for this application (shown in Figure 4(a)). Passing and lane change assistance may reduce or eliminate risks during these maneuvers, since they are often the source of serious accidents (shown in Figure 4(b)). Clearly these applications require close-range IVC with tight real-time constraints and can be implemented with either an SIVC (Sparse Roadside-to-Vehicle Communication) or a URVC (Ubiquitous Roadside-to-Vehicle Communication) system. Both systems can offer similar real-time guarantees and delays if properly designed, although an SIVC system may have a slight advantage as it faces reduced contention and direct links. These applications also have addressing based on ZOR; for example, immediately behind, in the right lane, or in the reverse direc- tion. Penetration directly influences the usability of these systems. Traveler information support applications provide updated local information, maps, and in general messages of relevance limited in space and/or time. These messages mainly focus on the local information and road warnings. Local information such as local updated maps, the location of gas stations, parking areas, and schedules of local museums can be downloaded from selected infrastructure places or from other “local” vehicles. Advertisements with, for example, gas or hamburger prices may be sent to approaching vehicles (shown in Figure 5(a)). Road warnings of many types (e.g., ice, oil, or water on the road, low bridges, or bumps) may easily be deployed by authorities simply by dropping a beacon in the relevant area (shown in Figure 5(b)). Figure 4. Traffic coordination and traffic assistance Figure 5. Traveler information support applications 235 Applications in Vehicular Ad Hoc Networks Figure 6. Communicate with different targets Only a few papers consider specific traveler information support applications. All information support applications require an SRVC system. IVC systems may augment the service provided by the SRVC, but cannot replace it. No special real-time requirements are necessary, and the penetration percentage has no impact on usability. A vehicle equipped with a vehicle communication system benefits from the applications independent of the OBU penetration rate. The addressing is once again based on relevance rather than individual vehicle IDs. The main focus of comfort applications is to make travel more pleasant. This class of applications may be motivated by the desire of passengers to communicate with either other vehicles or ground-based destinations such as Internet hosts or the public service telephone network (PSTN). Targeted vehicular communications allow localized communications (potentially multi-hop) between two vehicles (shown in Figure 6(a)). Voice, instant messaging, or similar communications may occur between the occupants of vehicle caravans traveling together for long distances, or between law enforcement vehicles and their “victims.” Note that this application does not scale to large network sizes. Vehicle to land-based destination communications (shown in Figure 6(b)) is arguably a very useful capability as it may enable an entire array 236 of applications, from email and media streaming to Web browsing and voice over IP. Unfortunately, land-based access requires a URVC system that may be prohibitively expensive in the near future. Finally, there are many other comfort applications. Tolls for roads and bridges can be collected automatically. Many nonstandard systems exist and work well. Parking payments can be made promptly and conveniently. Repair and maintenance records can be recorded at the garages performing them. Multimedia files such as DVDs, music, news, audiobooks, pre-recorded shows can be uploaded to the car’s entertainment system while the car is in the garage. TraFFic inFormaTion in VaneTs According to the definition of Hudson Valley Transportation Management Center (HVTMC) (“Hudson Valley Transportation Management Center (HVTMC),”), traffic information means dynamic information concerning about traffic. Thus, traffic information in VANETs means the dynamic information obtained from the on-board units in the vehicles. Compared to the travel information, like the maps of the streets or the maps of the bus routes, traffic information in VANETs is usually more immediate and local. As follow- Applications in Vehicular Ad Hoc Networks ing, we will introduce and discuss some reference papers about the collections, presentations, and applications of the data, such as immediate road traffic information and navigation information etc. Figure 7. Relevance of spatial-temporal speed surface street Traffic estimation Average velocity is the most commonly used parameter to show the traffic information, such as the real-time traffic network systems of cities and counties in Taiwan (“Real-Time Traffic Network Systems of Cities and Counties in Taiwan,”), all provide average velocity of road sections as an immediate reference. And the literature “Surface Street Traffic Estimation” (Noble et al., 2007) provides a simple way to determine the current road conditions by using data analysis through the GPS and the mobile communication systems. The approach is as follows: First of all, split the road section into small segments, and drive a car passing through the same route repeatedly to collect the road information (record the relationship between time and location). Secondly, update each GPS track data to the central server, and use the spatialtemporal traffic status plot to describe the traffic characteristics, and then obtained the temporal velocity- the average speed of the journey and the spatial velocity- the average speed of the section. Thirdly, they use Quadrant Clustering to classify the spatial-temporal traffic status plot to predict the present traffic condition by the classification of the data. Then present the traffic condition as good, normal (spatially good but temporally bad, or temporally good but spatially bad) and poor. And then transmit the information acquired through the internet. (Figure 7) Finally, transmit the packages through the internet. This method presents the road traffic situation with the effective spatial-temporal plot, taking the waiting time of traffic lights and intersections into account. estimating Velocity Fields on a Freeway from lowresolution Videos This is an approach using low-resolution video to estimate the velocity fields (Rice et al., 2006). Nowadays, there are many image processing techniques used in traffic surveillance (Kastinaki et al., 2003) . However this algorithm is not only used to estimate the speed for a single vehicle, nor to obtain the estimated value of the derivative through the intensity of each point, but rather to find out the most matching value by calculating the “intensity.” It is divided into two parts as intensity profiles and speed estimation. The first part, intensity profiles, is to choose four related fixed objects from the video (Figure 14.8) between the time interval [t – τ t + τ,], then mark them out with the frame, next find the best matching pattern in each frame, strengthen its intensity and use the centric coordinates to compute the projection matrix. Repeat the process. After some time, the intensity flow can be depicted so as to know the traffic conditions. In the second part, speed estimation, the basic idea is that the pattern of the intensity can present the speed of vehicles 237 Applications in Vehicular Ad Hoc Networks Figure 8. Filtered image and the corresponding intensity profile in the region. So the same pattern will be found at the next point in a short period. The processes are as follows: Fix the chosen frame, find out the intensify relationship d() between time instant t τk and t + τk, where located x - d and x + d. Then take Bayesian point to estimate the velocity and get the global vehicles’ velocity. local density estimation and dynamic Transmissionrange assignment (dTra) in Vehicular ad hoc networks In addition to speed, density is also a common parameter of traffic information. Here we will introduce a method, DTRA, in vehicular network which can estimate local density and dynamically adjust transmission radius without other vehicles or road side unit (RSU) (Artimy, 2007). According to the definition of traffic flow, the number of vehicles passing through a fixed-point per unit time could be computed as velocity (driving distance per unit time) product density (number of vehicles per unit distance). But as shown in Figure 14.9, traffic flow can not always reflect the traffic condition, such as conjunction or free 238 flow will show up at the same traffic flow. Thus, this paper provides a algorithm which only uses a single vehicle to estimate local density, and determine whether it is the case of congestion, then apply the result on the transmission range dynamically to meet the need of maintaining the high-connectivity of the network. Based on the two-fluid model (Robert et al., 1979), compute fraction (f) = the stopping time of test vehicle (Ts) divide the traveling time of the tested vehicle (Tt). And by NaSch-S2S model (Jost et al., 2003), they -1 h +1 é ù ê (1 - Ts / Tt ) ú + 1ú of get the density K = ê ê ú l¢ êë úû neighbor vehicles around, η and λ are given as the levels of the estimated traffic condition. At least, although the larger the radius of the inter-vehicle transmission range, the easier will be the communication; but on the same time, it will be easier to be intterrupted. Thus, the area of communiction can be adjusted according to the dynamic traffic condition mentioned above, reducing the radius of communication in high intensity condition and vice versa, so that highly connected network and data transmission rate can be achieved. Applications in Vehicular Ad Hoc Networks evaluation of a neighbor-based Vehicle density estimation scheme Most current methods about collecting and processing traffic information are still based on centralized communication model. It will be delivered to every vehicle after the processing of the central processing unit. However, this article is collected to predict the density of the vehicles on the roads based on the inter-vehicle communications (Panichpapibo et al., 2008). It takes two models to test, One-hop-Neifhbor Scheme and Two-hop-Neighbor-Scheme. The processes are as follows: (1) in the case of one-hop senario, vehicle launched to send “HELLO” packet to the vehicles which are in the communication range. And other vehicles send a message back to the collecting vehicle on receiving the packet. If the inter-vehicle space shows the exponential distribution, then the number of vehicles will show the poisson distribution (Wisitpongphan et al., 2007). At this time, the maximum probability of local density is the best expression to the global density estimation ( r̂ * ) equals to the numbers of neighbor vehicles (k) divide the communication radius of vehicles (z). (2) In two-hop situation, vehicle V broadcasts inquiring packets and receive responses from the vehicles within the communication range. Then send the packet to the furthest vehicle U in its communication range, and U become the probe vehicle doing the same work as vehicle V in the one-hop situation, collect the data along the direction and return the result back to vehicle V, finally the two-hop result is gotten. Evaluating the inaccuracy between estimation and actual density by MAE (mean absolute error), the authors find that the two-hop estimation is the better one. And the inaccuracy increases if the density is too low. The reason may due to the lack of vehicles in the communication range. In the end, the author suggests three ways to improve the estimation of the density. Firstly, enlarge the communication range to avoid the shortage of vehicles; secondly, rise the hops up; last and the most effect, use more samples for evaluation. estimating Travel Times and Vehicle Trajectories on Freeways using dual loop detectors There are many physical methods to depict traffic jams before 2002 (Nagatani, 2002). However, someone did not satisfied in using velocity and density as the only factor in the traffic information. This article provides a method to estimate travel time by using Dual Loop Detetor. The author uses the Dual Loop Detetor to record the speed of each vehicle and its arriving time at the Figure 9. Flow-density 239 Applications in Vehicular Ad Hoc Networks the certain point where the detector is set. Then draw chords to represent the directions and locations for vehicles. Because there may be free flow or conjunction at the same time, the situation of conjunction into consideration need to be taken (Figure 9), using the slop of conjunction uc to cut off the chords (Figure 10), then sum up the cutting chords to get the estimation of vehicles’ path and travel time. As the traffic information develops rapidly, the traffic information in VANETs is bound to become more important. Besides all the traffic informations mentioned above, which includes the vehicle speed, vehicle density, and travel time, there are still other traffic information; for example, the instant safety information. But no matter what kind of information is given, the most important aim is to let the users to arrive at their destinations safely and swiftly. Vehicle naVigaTion Technology In recent years, the expansion of cities scale and urban population increase as well as the development of automobile industry. Urban traffic problem such as traffic congestion, traffic accidents, transport environmental degradation, and energy shortages has become the world facing common problems. According to statistics, road congestion has caused a great amount of direct or indirect economic losses. Thus, in order to improve efficiency and security of transportation, every country has developed vehicle navigation and positioning system one after another. To solve the traffic congestion in a traditional way is to expand the roads or to construct new roads. But with the living environment nowadays, the urban population decreasing, the use of traditional methods is difficult to solve traffic congestion problems. Undoubtedly, the modern technology comes with the tide of fashion to solve the traffic problem effectively. It integrates computer technology, information technology, data communications 240 Figure 10. Travel time estimation “a” is a chord transmission technology, automation technology, artificial intelligence, and electronic technology etc. And establish a large-scale and all-round play transportation management system which is on time, accurate, and efficient. 14.4.1 Vehicle routing navigation system Navigation system is usually defined as the instrument which can be used as guidance for routing. From the early compass even later paper maps, the recent development of handheld GPS map navigation, on-board navigation systems (mainly can be divided into two types of portable and embedded), as well as PDA and Smartphone combined with navigation systems. Vehicle navigation system directs vehicle along a path from start position to the destination in the road network. The system attempts to generate an optimal path of guidance, almost all the systems try to find the shortest travel time for navigation. In other words, it attempts to search for the fastest route in the road network for guidance. The key influencing factors of the vehicle navigation system architecture are vehicle localization technology and the integrity of the electronic map. Global Positioning System is a widely used satellite localization system. There are 24 satellites distributed over the Earth. The satellite uses basic triangulation positioning principle, combined with the development of satellite and communication technology to GPS receiver to receive GPS satellite signals by the wave. It can accurately calculate Applications in Vehicular Ad Hoc Networks the time, position, and velocity of the area where are measured on the surface of the Earth at anytime and anywhere. Its advantages are global, high precision, real-time positioning speed, can be used for many users simultaneously, and free of charge. And the disadvantage is that it can not receive signals due to the signals obscured in some areas, such as behind buildings or tunnels, navigation signal interruption, therefore, must be again orientation. The components of vehicle navigation systems requirement: (1) positioning accuracy, combining different navigation technology to improve positioning accuracy. (2) Complete the correct geographic information systems, to update map data and road network instantly, and continued to increase the road property and other relevant information. (3) Friendly user interface. (4) Easily to operate, concern about the development of voice recognition capabilities, to provide voice-guided Service. (5) Price considerations. Information displayed can be divided into two categories, (1) Driving geographic information: to provide users with a wealth of traffic information, such as gas stations, parking area, restaurants, amusement parks, financial institutions etc. Most of the current systems are well build, and the content of the database is increase continuing. And the parts of the systems also provide users with custom landmark features. (2) Driving guidance information: to get the real-time traffic condition information, such as traffic congestions, traffic accidents, construction or temporary traffic control measures etc from communications technology. Users can avoid road congestion or accident location timely, and select one alternative path for their route planning. Based on the above classification, comment, and comparison of vehicle navigation system today (Schmitt et al., 2006), it will be divided into four main categories investigate separately for static and dynamic systems, deterministic and stochastic system, reactive and predictive system, as well as centralized and decentralized system into comparison. static vs. dynamic system One important distinction of routing guidance system is whether the system uses a static or a dynamic approach. This characteristic of the system dictates overall operation, no matter what vehicles receive and react to update path information during roadway propagation. A static routing algorithm determines and maintains a path for a user from the source to destination regardless of changing the traffic characteristics in time. In this system, vehicles already on the road with assigned paths do not react to any real-time network changes. Clearly this system represents an offline routing approach. Several papers investigate the idea of static path route selection such as in (Dijkstra, 1959; Dreyfus, 1969). On the other hand, dynamic path planning incorporates real-time traffic information and reacts to the changing conditions of the roadway. This allows updating of the path of a vehicle as the user proceeds to the destingtion. This type of routing provides a more robust solution than static routing for congestion alleviation. The differences between static and dynamic navigation system are system design and requirements. Static routing guidance systems require less roadway infrastructure and computing power, the system is vulnerable to events and congestion that occurs on the roadway. Dynamic routing guidance is a robust operation during events and congestion, but the system required more computing power and roadway infrastructure. Several papers investigate the idea of dynamic path route selection such as in (Hall, 1986; Karimi et al., 2004). deterministic vs. stochastic system The distinction between these two systems is mainly due to treating the cost of traveling from the 241 Applications in Vehicular Ad Hoc Networks origin to destination as a deterministic or stochastic cost. This cost is usually the travel time. Deterministic routing systems assume deterministic parameters for roadway links without regard to any random nature of traffic. These parameters may be updated frequently in time. The most commonly used metric for deterministic routing is the distance traversed by a vehicle along the route. Calculation of the path using a deterministic approach would yield the total distance. The total travel time is then calculated using mean vehicle speed without considering the influence of any variation due to the stochastic nature of the traffic. Several papers investigate the idea of deterministic path of routing selection such as in (Friesz et al., 1989; Zhan, 1997). Stochastic routing uses a more complex approach to vehicle routing in which the random nature of the traffic condition is also considered. In this approach, traffic along the road may be considered either as a stationary or non-stationary stochastic variable. Stationary stochastic routing guides vehicles along a path determined using only mathematical expectation of conditions deemed imporant by the algorithm. These conditions are assumed to be time invariant; therefore, they are not changing with time. In most cases of vehicle routing the mean and variance of travel time are used. Several papers investigate the idea of stationary stochastic path route selection such as in (Bertsekas et al, 1991; Alexopoulos, 1997). Nonstationary stochastic routing takes into account the changing characteristics of time with respect to the traffic condition. These changing conditions are familiar with mean travel time and the variance of the function which is time in one day. Several papers investigate the idea of stationary stochastic path route selection such as in (Seongmoon et al., 2005; Fu, 2001). Deterministic systems offer the simplest solutions while stochastic non-stationary solutions offers the most robust answer to the problem capturing the charateristics of traffic congestion, 242 at the cost of computational complexity. Stochastic routing systems clearly require a greater amount of data for routing guidance. Thus, the data requirement, especially historical data, increases computational time and data storage costs. reactive vs. predictive system This classification is very important. It controls the complexity, robustness, and vulnerability of events. The main differences between the two systems are their complexity and reliability. Reactive navigation routing is also a kind of feedback routing, which based on the recent situation of the travel network. The routing system determines the paths by real-time traffic information, but will not predict the coming situation. This system will provide the best route, which having least travel time, to the users. The system has a less complexity, but a greater vulnerability because of road congestion and unexpected events. Several papers investigate the idea of reactive path route selection such as in (Pavlis et al., 1999; Deflorio, 2003). Predictability routing of navigation can estimate the routes. It is based on the prediction models from the predicted situations of the future. The Predictive System using a repeated model of traffic information and useful historic information to predict the coming situations, which is based on linking or the region of network. The system will provide the best route with least travel time according to the estimated traffic conditions. Several papers investigate the idea of predictive path route selection such as in (Messmer et al., 1990; Kotsialos et al., 2002). Although the reactivity and the predictability routing are working in different ways, they both provide the best route to the users. We can say that the reactive system can be executed in the environment of lower cost and less complexity, and the prediction system will be able to avoid the generation of traffic congestion. Applications in Vehicular Ad Hoc Networks centralized vs. decentralized system Routing navigation can be dominated by a centralized system and monitoring dynamics of the entire network, or optimize the performance of a single user himself. Centralized system is focus on providing routing selections, and the distributed system provides personalized real-time road information selections. Centralized routing system uses a central unit to provide the routing selection. The system uses a predictive or reactive routing to guide vehicles for the best route to the destination in the network. Centralized routing is more complex than distributed system. This method provides more robust and reliable system to all hierarchies of the Internet. Centralized system will provide better network capacity to the network, the best of the road. It will calculate and deal with the road side infrastructures. Another concerned issue of the centralized system is the compliance rate of Internet users. If the instructions do not be followed by the drivers, the effectiveness of navigation systems will reduce while compliance rate decreasing. Several papers investigate the idea of centralized path route selection such as in (Yamashita et al., 2004 ; Tomkewitsch, 1991). Distributed routing navigation system required for individual users to determine all the routes of navigation, and each one can operate by the machine or man-made to decide the route of navigation. Road infrastructures can not control the selected routes of users, but it can provide current information and the parameters which has predicted of the network. The system may not be robust on the Internet hierarchy, because the optimization is for the situation of each user rather than to whole network. The critical difference of centralized and distributed routing navigation system is that the computation for users or infrastructures. Several papers investigate the idea of decentralized path route selection such as in (Wunderlich et al., 2000 ; Minciardi et al., 2001). In order to expand the commercialization of the product and to accelerate it into human life from technology products converted to daily necessities, related industries should study in market preference actively when technology developed. daTa disseminaTion in VaneTs data dissemination Data dissemination concerns about the transmission of information to intended receivers with certain design objectives. The design objectives that considered include low delay, high reliability, low memory occupancy, and low message passing overhead (Nadeem et al., 2006). Four services that have immediate application are unicast, multicast, anycast and scan. “Unicast with precise location” means a message should be delivered to node i in location l before time t . “Unicast with approximate location” means sending a message to node i before time t1 while that node was last known to be at location l with mobility m at time t2. “Multicast” means disseminating a message to all receivers in region r before time t. “Anycast” means disseminating a message to one among a set of possible destinations in region r before time t. “Scan” is to have a message traverse region r once before time t. Data delivery mechanisms define the rules for moving information through the network, for example “Node centric”, “Location centric”, “Opportunistic forwarding”, and “Trajectory based forwarding” .The “Node centric” approach specifies the routing path as a sequence of connected nodes. However the high vehicle mobility in V2V networks will quickly render inter-node connections invalid. The “Location centric” approach decouples the routing path from the intermediate nodes and the message is forwarded to the next hop(s) closer to the destination geographically. If a hole is encountered, efforts are made to find a path around it. “Opportunistic forwarding,” as 243 Applications in Vehicular Ad Hoc Networks suggested in, targets networks where an end-toend path cannot be assumed to exist. Messages are stored and forwarded as opportunities present themselves. When a message is forwarded to another node, a copy may remain with the original and be forwarded again later to improve reliability. “Trajectory based forwarding” directs messages along predefined trajectories. It was presented to work well in a dense network. Despite their sparseness, V2V networks should be a natural application of trajectory based forwarding because messages are moving along the “road” graph. mobility-centric data dissemination algorithm for Vehicular network MDDV is a “mobility centric” approach that combines opportunistic forwarding, geographical forwarding, and trajectory based forwarding. A forwarding trajectory is specified extending from the source to the destination (trajectory base forwarding), along which a message will be moved geographically closer to the destination (geographical forwarding). With an opportunistic forwarding approach, rules must be defined to determine which is eligible to forward a message, when a copy of the message should be passed to another vehicle, and when a vehicle should hold/ drop a message. They motivate the design by reference to a test scenario, geographical-temporal multicast. Geographical-temporal multicast is formally defined as: deliver a message to all vehicles in/entering region t before time t while the data source s is outside of r. Assume that a vehicle knows the road topology through a digital map and its own location in the road network via a GPS device. And suppose vehicles know the existence of their neighbors through some link level mechanism. But they do not assume a vehicle knows the location of its neighbors. The message dissemination information, e.g., source id, source location, generation time, destination region, expiration time and 244 forwarding trajectory, etc., is specified by the data source and is placed in the message header. A forwarding trajectory is specified as a path extending from the source to the destination region. One of the MDDV objectives is to deliver messages to their destination regions as soon as possible. A naive approach would be taking the path with the shortest distance from the source to the destination region. However, information propagation along a road depends largely on the vehicle traffic on it. A short road distance may not translate to short information propagation delay. High vehicle density often leads to fast information propagation. But vehicle traffic conditions change over time and vary from one road segment to another. Here, the authors only explore the static road network topology information since road networks are typically engineered to match transportation demands. They define d(A,B) as the “Dissemination length” of a road segment from road node A to B, which takes into account the static road information. d(A,B) = r(A,B) (m-(m-1)(ip+cjp)), 0< c < 1 where, r(A,B): road length from A to B i: number of lanes from A to B j: number of lanes from B to A But the traffic in the opposite direction of the desired information flow is less helpful than the traffic in the same direction of the information flow. Constant c is used to discount the opposite traffic flow. In the reference (Wu, Fujimoto, Guensler, & Hunter, 2004), when i =1, j =0, d(A,B)=r(A,B), then set m=5, p=0.1, c=0.005. The dissemination process consists of two phases: the forwarding phase and propagation phase. In the forwarding phase, the message is forwarded along the forwarding trajectory to the destination region. When the message reaches the Applications in Vehicular Ad Hoc Networks destination region, the propagation phase runs and the message is propagated to every vehicle in an area centered on the destination region before the message time terminated. During the forwarding phase, the authors call the “message head” that the message holder closest to the destination region along the forwarding trajectory. The vehicle taking the role of the message head may change over time as the message propagates. With perfect knowledge, every vehicle knows the message head vehicle in real time. When the message head tries to pass the message to other vehicles that may be closer to the destination region. During the propagation phase the message is propagated to vehicles without the message in the specified area. With the assumption that vehicles do not know the location of others, this is difficult to do. In both cases, the message is lost. To overcome this problem, they allow a group of vehicles near the real message head to actively forward the message instead of the message head vehicle only. The message head pair, the message head location and its generation time, is contained to the message. The “message head pair” provides the more correct knowledge of a message holder concerning the message head location. The actual message head can move either toward or away from the destination region along the forwarding trajectory within a short period of time. But it should move toward the destination region in the long term. For simplicity, the authors require that the message head location installed by a message holder never moves backward, which means that a message holder can only install a new message head location closer to the destination region than the one currently installed. To reduce the publication and dissemination of false information, only some vehicles are allowed to generate the message head pair. A message holder may assume either one of two roles: the message head candidate and non-message head candidate. Only a message head candidate can actively publish its current location as the message head location, and a non-message head candidate can only learn from received messages. There are rules for a message holder to transit between a message head candidate and nonmessage head candidate. Suppose the current time is tc, a vehicle’s current location is lc, and a vehicle’s installed message head pair is l, t where l is the message head location and t is the generation time. 1. 2. Non-message head candidate ⇒ message head candidate: During the forwarding phase, one important observation is that a vehicle passing its installed message location in a shorter period after the generation time is more likely to be the message head because after a long period the message may have already been forwarded far away toward the destination region along the trajectory. Thus, they stipulate that a non-message head candidate becomes a message head candidate if it passes its installed message head location toward the destination region before t + T1, where T1 is a system parameter. During the propagation phase, message holders moving into the destination region assume the role of the message head candidate. Message head candidate ⇒ non-message head candidate: During the forwarding phase, there are two transition rules: 1. If the message head candidate leaves the trajectory or moves away from the destination region along the trajectory, it becomes a non-message head candidate; 2. If a message head candidate moves toward the destination region along the trajectory, it stays as a message head candidate until it receives the same message with another message head pair l n , tn where ln is closer to the destination region than lc. 245 Applications in Vehicular Ad Hoc Networks During the propagation phase, a message head candidate becomes a non-message head candidate once it moves out of the destination region. A message holder updates its installed message head pair with the information from received messages. Two messages differing only in the message head pair are two versions of the same message. One message version with message head pair li , ti is said to be newer than another message version with message head pair l j , t j 1. 2. li is closer to the destination region than lj li = lj but ti > tj Data exchange can be triggered by several types of events. In MDDV, data exchange is triggered by: new messages, newer message versions or older message versions are received, or new neighbors appear. Transmissions triggered by new messages or newer message versions serve to quickly propagate messages or dissemination status. Transmissions triggered by older message versions can help eliminate false/obsolete information. This scheme has both the advantages of fast delivery and high delivery reliability. It is called the “full protocol.” The data exchange algorithm is defined as: 1. 246 Forwarding phase: A message holder can be either one of following dissemination states: the active state, the passive state, and not eligible to transmit at all. A message holder in the active state runs the full protocol to actively propagate the message while a message holder in the passive state only transmits the message if it hears some older message version. The active propagation can help populate the message, move the message closer to the destination region or update dissemination status. The passive updating serves to eliminate false/obsolete information only. In the active state, tc < t + T2, and lc between L2 with l. In the passive 2. state, tc < t + T3, and lc between L3 with l. T2, T3, L2, L3 are system parameters and T2 < T3, L2 < L3. Propagation phase: A message holder can be either in the active state or not eligible to transmit. A message holder is in the active state if tc < t + T2 and lc is within the distance L2 from l. An opportunistic forwarding mechanism must determine when to store/drop a message. The design decision can affect delivery reliability, memory usage and message overhead. The decision to store/drop messages can be based on a vehicle’s knowledge of its future movement trajectory. For example, assume vehicles are aware of its own near future movement trajectory, a message holder may decide to drop a message if it knows that continually holding the message can no longer contribute to suppress unnecessary message transmissions based on its future movement trajectory. In MDDV every vehicle stores whatever it overhears since this is almost free except occupying memory buffers. A vehicle drops a message when the vehicle leaves the passive state during the forwarding phase, leaves the active state during the propagation phase or the message expiration time elapses. concept of replication In the Intelligent Transportation System (ITS), the car to car wireless network and car’s communication to the infrastructure (C2X communication) have been given high priority recently. This trend is corresponding to use basis of the C2X active safety applications to reduce the traffic accident or other seriously expected. The networks constructed by C2X is called “Vehicular Ad Hoc Networks (VANETS)”, it runs distributed approach and does not depend on the existing infrastructure to operate. Because of the C2X node is generally moving, so the whole Applications in Vehicular Ad Hoc Networks network is divided by a number of Ad Hoc network composed. Due to the region information providing an appropriate way to describe that region, its applications have various C2X’s characteristics. For example, in the case of traffic jam alert, the main of provision will directly affect the situation of information to the vehicles, as vehicles have nearly the same lane. Other major factors include the communication between vehicles and infrastructure. To imagine a case, if the traffic light can broadcast the current signal to the nearest car. In this case where the vehicles close to the traffic signal, it can specify the affected roads and driving directions. However, this assignment is hard to achieve in the reality traffic situation, because the moving nodes may already get messages in hundred meters away and roads are generally not linear. Therefore, to estimate the information received by vehicles is a challenging problem. This article introduce about how to deal with the roads affected that allow changing and ensure the problem that vehicles can receive the relevant information. replication concept of regional road network The approaches in real traffic environment describing the region are often unrealistic, because it ignore the specific situation of the driving environment. In the affected region may be more complicated to operate and not at all the nodes in this area will participate in. In fact, the relationship about the car direction is an important problem. The issue of concept is to use some of the challenges faced the basic position of a temporary mechanism to change the essential part of the street. This means that each vehicle contains C2X are required a wireless communication components and similar Global Position System (GPS) location systems. However, not every vehicle requires Global Position System. This method keeps storing their current and recent location information in a period of time. Therefore vehicle can accurate to provide their tracking over the past. It can be every kind of straight-line (ex: driveway) or curved-line (ex: driveway exit or highway). If the slippery road warning occurs, it is necessary to send the warning message to the vehicle. Thus, the position information of that area must be stored actually. The saving points are probably in hundred meters away. All settings can be illustrated that greatly reduce digital map by the regional road network in the travel path of vehicles to access. To decide the location of information storage and the use of the map are very easy problems, because most of the locations are dependent on the technology of current trends (Roessler et al., 2006). Therefore, it contains the following features: • • Form of the Roads: The form of different roads will be different number of supporting points. If there is lots of curve in the road, it is necessary to rebuild the storage location information. Otherwise, it only needs two points of information while the road is straight. Velocity of the Vehicle: When vehicle moves with high speed in a short time interval, it generally store high change location information. To go along with the interpolation points, the width of the addressing region (corridor), is needed. This width can be equal to the whole set of supporting points in order to reduce the amount of data. However, it might also be necessary to specify different width parameters in case of a complicated road form. Figure 11 shows the example that is from X1/Y1,…,Xn / Yn of supporting points. It can be found that the distance between two points are different from the whole addressing region. To collect this kind of additional information from supporting points, i.e. the expected 247 Applications in Vehicular Ad Hoc Networks Figure 11. Example for concept of replication information of driving directions which is form the addressing area linked to a dedicated road section. This new method includes several advantages that are summarized below: • • • • 248 Low Basic Hardware Necessary: Whole information needed like current position, driving direction and time can be got from a basic positioning system. However, the memory demand can be constraint by limitation of the maximum number of cached position data. Simple Relevance Evaluation: Using the received information vehicle to make a first evaluation of the relationship of the information even if they are not equipped with a navigation system. Scalability: There are several method to improve. For example, if a vehicle is equipped a navigation system, it may also solve the problem that address streets were not on its recent route. But it might also be affected. Flexible: This approach can apply to any mobile and fixed nodes. If an immobile station like a traffic sign is the sender, the position information may alternatively be hard-coded in the device. As mentioned above, the addressing approach is not restricted to the vehicle to vehicle scenario that has been used to describe the technique. It can be further applied in different C2X-applications like the communication between a car and a traffic signal system. reFerences Alexopoulos, C. (1997). State Space Partitioning Methods for Stochastic Shortest Path Problems. Networks, 1(30), 9–21. doi:10.1002/(SICI)10970037(199708)30:1<9::AID-NET2>3.0.CO;2-H Artimy, M. (2007). Local Density Estimation and Dynamic Transmission-Range Assignment in Vehicular Ad hoc Networks. IEEE Transactions on ITS, 18(3), 400–412. Applications in Vehicular Ad Hoc Networks Bertsekas, D. P., & Tsitsiklis, J. N. (1991). An Analysis of Stochastic Shortest Path Problems. Mathematics of Operations Research, 16(3), 580–595. doi:10.1287/moor.16.3.580 Deflorio, F. P. (2003). Evaluation of a reactive dynamic route guidance strategy. Journal of the Transportation Research Board Transportation, 11(5), 375–388. Dijkstra, E. W. (1959). A note on two problems in connection with graphs. Nuerische Mathematik, 1, 269–271. doi:10.1007/BF01386390 Dreyfus, S. E. (1969). An Appraisal of Some Shortest Path Algorithms. Operations Research, 17, 395–412. doi:10.1287/opre.17.3.395 Friesz, T. L., Luque, J., Tobin, R. L., & ByungWook, W. (1989). Dynamic network traffic assignment considered as a continuous time optimal control problem. Operations Research, 37(6), 893–901. doi:10.1287/opre.37.6.893 Fu, L. (2001). Adaptive Routing Algorithm for In-Vehicle Route Guidance Systems with RealTime Information. Transportation Research Part B: Methodological, 35(8), 749–765. doi:10.1016/ S0191-2615(00)00019-9 Hall, R. (1986). The Fastest Path through a Network with Random Time- Dependent Travel Time. Transportation Science, 20(3), 182–188. doi:10.1287/trsc.20.3.182 Hudson Valley Transportation Management Center (HVTMC) (n.d.). Retrieved from http://www. hudsonvalleytraveler.com/Glossary.html Jost, D., & Nagel, K. (2003). Probabilistic traffic flow breakdown in stochastic car following models. Transportation Research Record, 1852, 152–158. doi:10.3141/1852-19 Karimi, A., Hegyi, A., Schutter, B. D., Hellendoorn, J., & Middelham, F. (2004, Oct.). Integrated model predictive control of dynamic route guidance information systems and ramp metering. Paper presented at the Proceedings of the 7th International IEEE Conference on Intelligent Transportation Systems (ITSC 2004), Washington, DC. Kastinaki, V., Zervakis, V., & Kalaitzakis, K. (2003). A survey of video processing techniques for traffic applications. Image and Vision Computing, 21(4), 359–381. doi:10.1016/S02628856(03)00004-0 Kotsialos, A., Papageorgiou, M., Diakaki, C., Pavlis, Y., & Middelham, F. (2002). Traffic flow modeling of large-scale motorway networks using the macroscopic modeling tool METANET. Journal of IEEE Transaction on Intelligent Transportation Systems, 3(4), 282–292. doi:10.1109/ TITS.2002.806804 Messmer, & Papageorgiou, M. (1990). METANET: A macroscopic simulation program for motorway networks. Journal of Traffic Engineering & Control, 31(8-9), 466-470. Minciardi, R., & Gaetani, F. (2001). A decentralized optimal control scheme for route guidance in urban road networks. Paper presented at the Proceeding of IEEE Intelligent Transportation Systems, Oakland, CA. Nadeem, T., Shankarand, P., & Iftode, L. (2006, July). A comparative study of data dissemination models for vanets. Paper presented at the Proceeding of 3rd Annual International Conference on Mobile and Ubiquitous Systems (MOBIQUITOUS), San Jose, CA. Nagatani, T. (2002). The physics of traffic jams. Reports on Progress in Physics, 65(9), 1331–1386. doi:10.1088/0034-4885/65/9/203 249 Applications in Vehicular Ad Hoc Networks Noble, B., Yoon, J., & Liu, M. (2007, June 11-14). Surface Street Traffic Estimation. Paper presented at the Proceedings of International conference on Mobile systems, San Juan, Puerto Rico. Panichpapibo, S., & Pattara-atikom, W. (2008, Oct.). Evaluation of a neighbor-based vehicle density estimation scheme. Paper presented at the ITST 2008. 8th International Conference, Phuket. Pavlis, Y., & Papageorgiou, M. (1999). Simple Decentralized Feedback Strategies for Route Guidance in Traffic Networks. Transportation Science, 33(3), 264–278. doi:10.1287/trsc.33.3.264 Real-Time Traffic Network Systems of Cities and Counties in Taiwan. (n.d.). Retrieved from http:// www.e-traffic.com.tw Rice, J., & Cho, Y. (2006). Estimating Velocity Fields on a Freeway From Low-Resolution Videos. IEEE Transactions on Intelligent Transportation Systems, 7(4), 463–469. doi:10.1109/ TITS.2006.883934 Robert, H., & Ilya, P. (1979). A Two-Fluid Approach to Town Traffic. Science, 204(4389), 148–151. doi:10.1126/science.204.4389.148 Robinson, C. L., Caveney, D., Laberteaux, K., & Caminiti, L. (2006, September 29). Efficient Coordination and Transmission of Data for Cooperative Vehicular Safety Applications. Paper presented at the Proceedings of the 3rd international workshop on Vehicular ad hoc networks, Los Angeles, California, USA. Roessler, B., & Bruns, C. (2006, Mar.). Replication of Local Road Networks for Regional Information Dissemination in VANETS. Paper presented at the Proceedings of the 3rd Workshop on positioning, Navigation And Communication (WPNC’06), University of Hannover, Germany. 250 Ryu, B., Yeung, G., Krishnan, H., Yin, J., ElBatt, T., Habermas, S., et al. (2004). Performance Evaluation of Safety Applications over DSRC Vehicular Ad Hoc Networks. Paper presented at the Proceedings of the 1st ACM international workshop on Vehicular ad hoc networks, Philadelphia, PA. Schmitt, E. J., & Jula, H. (2006, Sep.). Vehicle Route Guidance Systems: classification and Comparison. Paper presented at the Proceedings of the IEEE Intelligent Transportation Systems Conference (ITSC2006), Toronto. Seongmoon, K., & Lewis, M. E., & III, C. C. W. (2005). Optimal vehicle routing with real-time traffic information. IEEE Transactions on Intelligent Transportation Systems, 6(2), 178–188. doi:10.1109/TITS.2005.848362 Sichitiu, M. L., & Kihl, M. (2008). Inter-Vehicle Communication Systems: A Survey. IEEE Communications Surveys & Tutorials, 10(2), 88–105. doi:10.1109/COMST.2008.4564481 Taleb, T., Sakhaee, E., Jamalipour, A., Hashimoto, K., Kato, N., & Nemoto, Y. (2007). A Stable Routing Protocol to Support ITS Services in VANET Networks. IEEE Transactions on Vehicular Technology, 56(6), 3337–3347. doi:10.1109/ TVT.2007.906873 Tomkewitsch, R. v. (1991). Dynamic route guidance and interactive transport management with ALI-SCOUT. Journal of IEEE Transactions on Vehicular Technology, 40(1), 45–50. doi:10.1109/25.69971 Wisitpongphan, N., Bai, F., Mudalige, P., Sadekar, V., & Tonguz, O. (2007). Routing in sparse vehicular ad hoc wireless networks. IEEE Journal on Selected Areas in Communications, 25(8), 1538–1556. doi:10.1109/JSAC.2007.071005 Applications in Vehicular Ad Hoc Networks Wu, H., Fujimoto, R., Guensler, R., & Hunter, M. (2004, Oct.). Mddv: a mobility-centric data dissemination algorithm for vehicular networks. Paper presented at the Proceedings of ACM Vehicular ad hoc networks (VANET), Philadelphia, PA. Yamashita, T., Izumi, K., & Kurumatani, K. (2004). Car navigation route information sharing for improvement of traffic efficiency. Paper presented at the Proceeding of 7th international IEEE conference on Intelligent Transportation Systems, Washington, D.C., USA. Wunderlich, K. E., Kaufman, D. E., & Smith, R. L. (2000). Link travel time prediction for decentralized route guidance architectures. Journal of IEEE Transaction on Intelligent Transportation Systems, 1(1), 4–14. doi:10.1109/6979.869017 Zhan, F. B. (1997). Three fastest shortest path algorithms on real road network: data structure and procedures. Journal of geographic information and decision analysis, 1(1), 69-82. 251 252 Chapter 15 DTN Technologies for Vehicular Networks Kun-Chan Lan National Cheng Kung University, Tainan, Taiwan, R.O.C. absTracT A Delay Tolerant Network (DTN) is one type of challenged network where network contacts are intermittent or link performance is highly variable or extreme. In such a network, a complete path does not exist from source to destination for most of the time. In addition, the path can be highly unstable and may change or break unexpectedly. To make communication possible in a delay tolerant network, the intermediate nodes need to take custody of data during the blackout and forward it toward the destination when the connectivity resumes. A vehicular network nicely falls into the context of DTN since the mobility of vehicles constantly causes the disruption of link connectivity’s between vehicles. In this chapter, the authors discuss some research challenges and issues which might occur in a Delay Tolerant Network and how they are related to vehicular networks. inTroducTion Delay/disruption tolerant networks (DTN) can be used to increase the robustness of the network where the network protocols must be explicitly designed to perform despite frequent disruptions. A typical DTN consists of a set of wireless nodes, some or all of which are mobile. The nodes can range from small sensor nodes, to mid-sized devices carried by people, robots or vehicles, to larger semi-permanent DOI: 10.4018/978-1-60566-840-6.ch015 installations. Due to node mobility or limited radio range, end-to-end paths do not always exist between some or all the nodes. As a result, data are transferred in a store-carry and forward paradigm. Vehicular ad hoc networks (VANET) have been envisioned to be useful in road safety and many commercial applications. For example, a vehicular network can be used to alert drivers to potential traffic jams, providing increased convenience and efficiency. It can also be used to propagate emergency warning to drivers behind a vehicle (or incident) to avoid multi-car collisions. To realize this vision, the FCC Copyright © 2010, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global is prohibited. DTN Technologies for Vehicular Networks has allocated 75 MHz of spectrum for dedicated short range communications (vehicle-vehicle or vehicle-roadside), and the IEEE is working on standard specifications for inter-vehicle communication. As technology advances make it more feasible and cost-effective to produce vehicles that are equipped with communication capabilities that allow for inter-vehicle communication, large scale vehicular ad hoc networks are expected to be available in the near future. There are various kinds of VANETs based on the entities involved, such as vehicle to vehicle communication (V2V), vehicle to roadside communication (V2R) or vehicle infrastructure integration (VII), and roadside to roadside communication (R2R). There are different ways to look at DTN and VANET. At one extreme, DTN is more general, and VANET is a special kind of DTN. At another extreme, DTN techniques are only used in some VANET applications when vehicles are far away from each other, but not in safety related applications which have strict delay requirements. Nevertheless, DTN and VANET have many common characteristics which present challenges and opportunities for the research community. Many unique characteristics of VANET bring out new research challenges. First, due to fast vehicle movement, network topology and channel conditions change rapidly. As a result, many well-studied structures such as tree, clustering, grid, are extremely hard to set up and maintain. Second, the network density is highly dynamic. The traffic load is low in rural areas and during night, which may result in frequent disconnections and network partitions. On the other hand, during rush hours or traffic congestion, the network density is very high, which may generate data collisions and result in network congestion. Third, the vehicle mobility is partially predictable since it is limited by the traffic pattern and the road layout. These unique characteristics bring out research issues at different layers of the network stack. At the physical layer, directional antenna/ MIMO techniques may be applied to increase the network capacity. At the data link layer, new MAC protocols should be designed to meet latency and reliability requirements, especially for safety related applications. Because vehicles move along roads, directional-antenna-based MAC mechanisms might be especially useful for VANET. At the network layer, protocols should be designed to exploit the mobility to maintain the route. At the transport layer, new protocols should be designed to tolerate routing layer disruptions. From the network perspective, security and scalability are two significant challenges. Although power efficiency in VANET is less of a concern, scalability may still be critical. Due to the nature of the vehicular applications, there might be more flooding/broadcasting in VANET than in traditional ad hoc networks. This could easily create network congestion if the communication protocols are not welldesigned. There are many challenges and opportunities in DTNs. Many VANET applications can be built on DTNs, and hence most of the challenges and opportunities presented in VANET are also valid in DTNs. As social networking becomes popular, more applications can be built over DTNs. Also, a number of non-interactive applications related to sensor networks can be adapted to DTNs. Unlike Internet applications, DTNs provide an opportunity for aggregation, replication, and innetwork processing as data traverses the network. Traditionally, DTN is treated as an edge solution for the future Internet paradigm. It is a challenge to re-think the Internet as a DTN. Further, DTNs may rely on some level of infrastructure support to improve their performance, and hence hybrid DTN/infrastructure networks are another area to explore. There are many technical issues to explore in DTNs. First, how to achieve service discovery, context awareness and group communication, which are standard modules in distributed sys- 253 DTN Technologies for Vehicular Networks tems. Due to inherent delays in DTNs, these popular services need to be redesigned. Second, we should define the communication paradigm in DTN. In traditional networks, the end-to-end concept is used. In DTN, many issues need further investigation. For example, what are the network interfaces? Does DTN have a flow concept? Third, DTN is based on the idea of store-carry and forward. Although forwarding has been well studied in traditional networks, we don’t have enough knowledge on the role of storage and it is not clear how to manage storage in DTNs. Finally, there are many other research issues such as security, the role of network coding for DTNs, and network characterization (based on e.g. Graph Theory). One key feature of VANET/DTN is the mobility model, which is very important for future research in this area. In VANET, the mobility model can be based on previous work done by researchers in other fields such as civil engineering. However, more work should be done on porting these results to our community. This can be in the form of adding mobility models for VANET to well known simulation platforms such as ns2. For general DTN, the mobility model is not that clear, and is worth further investigation. Further, communication and data traffic models may be application-specific. In summary, VANET/DTN are well outside traditional networking assumptions and challenge our thinking about future Internet designs. In this chapter, we present an overview of existing research activities in DTN. dTn A Delay Tolerant Network (DTN) is a developing system that suffers frequent and long duration network partitions. It is an overlay on top of multiple network partitions, which may be challenged by limitations such as intermittent and possibly unpredictable loss of connectivity, 254 long or variable delay and high error rates. For the intermittent connectivity, an end-to-end path between the source and the destination may not exist. Therefore, TCP/IP protocol will break in this kind of environment because an end-to-end path between source and destination may only exist for brief and unpredictable periods of time. Furthermore, in DTN, there are long propagation and variable queuing delays. Many Internet protocols which are designed assuming quick return of acknowledgements and data fail to work in such networks. The problems of DTN as mentioned above are addressed by adopting store-carryforward message switching. Here entire chunks of message are transferred from one storage place to a storage place in another node along a path that is expected to reach the destination. There are many applications for DTN. For particular field, this kind of applications is used in the special environment that is to be tolerant of long delay and high error rate. The Interplanetary Internet project must encompass both terrestrial and interplanetary links. Sami Network Connectivity (SNC) Project (Lindgren et al., 2007) focuses on establishing Internet communication for Sami population of reindeer herders who live in remote areas. DARPA (Krotkov, et al., 1999) is researching and developing the capability to perform urban reconnaissance with teams of small, low-cost, semiautonomous mobile robots. For developing regions networks, the networks of developing countries or the outlying area are not usually perfect, so the foreign networks of a lot of areas are being developed, even there is no network at all. In (Brewer et al., 2005), several applications are developing for this kind of region from education to health care and government services. This project is supported by (Tier, 2009). For environmental monitoring, this application is to monitor the animals or environment. In (The zebranet wildlife tracker, 2009), this project is to track the Wildlife Tracker. DTN Technologies for Vehicular Networks Figure 1. The DTN protocol stack or more DTN regions and may optionally be a host, so it must have persistent storage and support custody transfers. challenges an oVerView oF delay ToleranT neTworks architecture In a DTN, a network is typically separated into several network partitions which called regions. Traditional applications are not suitable for this kind of environment because they normally assume that the end-to-end connection must exist from source to destination. The DTN enables the devices in different regions to interconnect by operating message in store-carry-forward message fashion. The intermediate nodes implement store-carry-forward message switching mechanism by overlaying a new protocol layer, called the bundle layer (Scott et al., 2007 ; Segui et al., 2006). The DTN protocol stack is as shown in Figure 15.1. In a DTN, each node is an entity with a bundle layer which can act a host, router or gateway. When node acts a router, the bundle layer of router requires persistent storage to queue bundles and forwards the bundles within a single region. On the other hand, the bundle layer of gateway is not only provision the store-carryforward mechanism, but provision the conversion between the lower-layer protocols of the regions they span, as shown in Figure 15.1. In other words, a DTN gateway can forward bundles between two In a DTN, when nodes move away or turn off their power to conserve energy, links may be disrupted or shut down periodically. These events result in intermittent connectivity. When no paths exist between source and destination, the network partition occurs. Therefore, nodes need to communicate with each other via opportunistic contacts through store-carry-forward operation. In this section, we consider three specific challenges in an opportunistic network: the contact opportunity, the node storage and the power conserving. contact Due to the node mobility or the dynamics of wireless channel, a node might make contact with other nodes at an unpredicted time. Since contacts between nodes are hardly predictable, they must exploited opportunistically for exchanging messages between some nodes that can move between remote fragments of the network. Burns et. al. (2005) classified the routing methods for DTN based on characteristics of participants’ movement patterns. The patterns are classified according to two independent properties: their inherent structure and their adaptiveness to the demand in the network. Other approaches propose message ferries to provide communication service for nodes in the deployment areas (Zhao et al., 2004, May ; Zhang et al., 2007 ; Chen et al., 2007 ; Tariq et al., 2006). In addition, the contact capacity needs to be considered. In other words, how much data can be transferred between two nodes when two nodes contact? Hui et. al. (Chaintreau et al., 2005) defined two parameters, contact duration and inter-contact time, that are the important factors in determining the capacity of the networks. 255 DTN Technologies for Vehicular Networks storage As the described above, to avoid dropping packets, the intermediate nodes are require to have enough storage to store all messages for unpredictable periods of time until next contact. In other words, the required storage space increases a function of the number of messages in the network. Therefore, the routing and replication strategies must take the storage constraint into consideration. Vahdat and Becker (Vahdat et al., 2000) used Epidemic Routing by flooding the network to exploit the best possible delivery delay brought by mobility. This scheme achieves the optimal delay with unlimited relay buffers. However, such a multiple-copy scheme generally incurs significant overhead on storage constraint. Ip et. al. (2007) proposed a buffer-management strategy, RRFS-with-RandomDrop, to avoid head-of- line blocking in the FIFO case. They showed that the proposed strategy can reduce the degradation of average delivery delay performance. Device Energy During opportunistic contacts, the nodes may need to communicate in order to transfer the data to destination. For conserving the device energy, the node may need to know the contact time and the duration for transfer the complete or segmented bundle in order to wake up at the contact time. Jun et. al. (2005) proposed a power management in DTN depending on knowledge-based mechanisms similar to the DTN routing algorithms (Jain et al., 2004). Two extremes of knowledge are complete knowledge and zero knowledge. The former has the precise time of contacts, so it can provide the upper bound on performance. The latter do not has any knowledge, so a node chooses a random time and broadcasts a beacon. If it does not receives a beacon, the node sleeps and wakes up at the beginning of the next beacon time. In the partial knowledge, a node knows statistics some information about the contact durations and waiting times 256 between contacts with other nodes. Depending on the statistical information, a node may expect to the waiting time before the contact. However, those methods may be applied to the network that followed a regular schedule. In (Jun et al., 2006), the authors propose a hierarchical power management in DTN. The energy in the architecture can be conserved by using lower-power, long-range, radio to discover the contacts and then wake the high-power, short-range, radio to prepare the data transmission. However, these architecture may be inefficient because mobile node may not enter into high-power radio range and do not reduce the cost of other idle hardware components. Similar to (Jun et al., 2006), a throwboxs with power management is proposed in (Banerjee et al., 2007) . In the architecture, the mobile node beacon their position, direction and speed using the long-range radio when it moves to the range of tier-0, long-range radio. If the throwbox hears a beacon, it predicts how long the mobile node will contact with the tier-1, short-range radio. By predicting the trajectory of an oncoming mobile node, the tier-0 wakes the tier-1 to undertake the data transmission. physical layer In order to operate under the challenged network (Fall et al., 2003), the concept of long-lived is an important requirement. For some applications, i.e. underwater sensor network, terrestrial sensor network, they must design the suitable communication hardware. Existing underwater communication systems have some useful physical design for DTN. Freitag et al. (2005) have developed a compact and low-power acoustic modem, called Micro-modem. The low-power design is an enabling factor for long-lived sensor networks over battery-powered nodes. In (Wills et al., 2006), authors propose an inexpensive and low-power hardware modem to save the power in order to provide long-lived operator. For low- DTN Technologies for Vehicular Networks power operation, the power-saving is to use the wakeup tone receiver to trigger the more expensive data receiver. Authors also assume the radio will be duty cycled, alternating on-and-off periods frequently when active, and with long-term off periods when inactive. mac layer As the section III, the medium access control (MAC) protocol needs to be modified for the high latency requirement. The communicating nodes in DTN may suffer from high latency problem. Similarly, the underwater acoustic communication may also suffer from latencies larger than radio communication. In (Rodoplu et al., 2005), authors extend S-MAC’s schedule synchronization to sender-receiver pairs in underwater. This protocol explains how to achieve a locally synchronized schedule even in the presence of long propagation delays. Each node schedules the time to transmit the next packet, and broadcasts this information by attaching it to the current data packet. While hearing the broadcasts, the other nodes will know when to wake up for the subsequent packet. However, in order to operate at a low collision rate, each node requires a small duty cycle, which makes throughput low. neTwork layer In this section, we discuss some routing solutions for a DTN. Based on the number of copies of a message forwarded by the node, we can define two different routing schemes: forwarding-based (single copy) approach and flooding-based (multiple copies) approach. In forwarding-based approach, there is only one single custodian for each message to help forwarding the message to destination. When the current custodian forwards the copy to an appropriate next-hop neighbor, this neighbor becomes the message’s new custodian. The same process is repeated until the message finally reaches its destination. This approach tries to reduce the buffer usages and the number of message transferred in the network. But it may suffer the long delays and low delivery ratios. On the other hand, flooding-based approach may generate multiple copies of the same message. Each message can be routed independently for increased efficiency and robustness. This approach achieves lower delays and higher delivery ratio at the cost of a larger buffer space and more message transfers. Forwarding g-based approach In the forwarding-based scheme, based on what type of knowledge nodes use to select the appropriate or the best path to destination node, the prior studies can be classified into three categories: direct transmission, location-based, Knowledgebased and control-movement based. Direct Transmission Spyropoulos et. al. (2004) is proposed a simple single-copy routing called direct transmission routing. In this approach, after the source node generates a message, the message is hold by the source node until it reaches the destination node. The main advantage of this scheme is that it incurs minimum data transfers for message deliveries. On the other hand, although having minimal overhead, this scheme may incur very long delays for message delivery since the delivery delay for this scheme is unbounded (Grossglauser et al., 2002) . Location-Based In the location-based approach, nodes will choose the neighbors who are closest to the destination to pass the message. LeBrun et al. (2005) proposed a method using the motion vector (MoVe) of mobile nodes to predict their future location. The MoVe 257 DTN Technologies for Vehicular Networks scheme uses the knowledge of relative velocities of a node and its neighboring nodes to predict the closest distance between two nodes. After the nodes’ future locations are calculated, messages are passed to nodes that are moving closer to the destination. As compared to epidemic routing, this approach has less control packet overhead and buffer usage. Leguay et al. (2006) presented a strategy that uses a virtual coordinate routing called mobility pattern spaces (MobySpace). In this approach, the node coordinates are composed of a set of dimensions, each dimension represents the probability that a node will be found in a specific location that is a virtual expression of the mobility pattern and does not geographic coordinate of the node. There are various destination functions are computed on this vector. They showed that this approach consumes less resource than epidemic routing when they deliver large amount of bundles. Knowledge-Based In the knowledge-based approaches, based on certain knowledge about the network, the source and intermediate nodes decide which nodes to forward messages as well as whether it should transmit the message immediately or hold the message until it meets a better node. Jain et al. (2004) proposed the knowledge-based routing scheme which is the first study in this area. Depending on the amount of knowledge about network topology characteristics and traffic demand, they define four knowledge oracles. Each oracle presents some particular knowledge of network. Based on the available oracles, the authors present a corresponding routing algorithm. The first algorithm is called Minimum Expected Delay (MED). In this algorithm, when the contact summary oracle, which contains information about aggregate statistics of the contacts, is available, Dijkstra with time-invariant edge costs based on average waiting time is used to find the best route. The second algorithm is called Earliest Delivery 258 (ED) where the contact oracle is available. The contacts oracle contains information about contacts between two nodes at any point in time. When this algorithm assumes the queuing delay is zero, the modified Dijkstra with time-varying cost function based on waiting time is used to find the route. The third algorithm is Earliest Delivery with Local Queuing (EDLQ), which uses the local queue occupancy to add an estimate of the queuing delay to the ED algorithm. Final algorithm is the Earliest Delivery with All Queues (EDAQ) where the contact oracle and queuing oracle is available. The queuing oracle gives information about instantaneous buffer occupancies (queuing) at any node at any time. This algorithm adds the queuing oracle to ED. This approach assumes that the accurate information about the oracle is known in advance. This assumption may be workable for some scenarios where the node movement is predictable, i.e., city bus. Musolesi et al. (2005) present the ContextAware Routing (CAR) protocol that provides an asynchronous communication for message delivery. In a DTN, since the receiver is often not in the same connected network, synchronous delivery of message is typically not possible. In CAR, if a message cannot be delivered synchronously, the message is sent to a host that has the highest probability of successful delivery and acts as a message carrier. The delivery probability process is based on the evaluation and prediction of context information using kalman filters. The prediction process is used during temporary disconnection and the process is continued until it is possible to guarantee certain accuracy. In addition, the epidemic routing can be considered optimal in terms of delivery ratio because each message is propagated to all accessible nodes which have large buffers to hold the messages. They showed in their simulations that if the buffer size is small, the packet delivery ratio of CAR is batter than the packet delivery ratio of epidemic routing due to that CAR only creates a single copy for each message. DTN Technologies for Vehicular Networks Burgess et al. (2006) proposed an effective routing protocol called MaxProp. A node uses MaxProp to schedule packets transmission to its peers and determines which packets should be deleted when buffer space is almost full. Packets are scheduled based on the path likelihoods to peers according to historical data. In addition, several complementary mechanisms, including acknowledgments, a head-start for new packets, and lists of previous intermediaries are used in this approach. They showed that their approach performs better than the protocols that have access to an oracle (Jain et al., 2004) that knows the schedule of meetings between peers. Kun et al. (Tan et al., 2003) proposed a shortest expected path routing (SEPR) similar to link-state routing to maintain a topology map to each other. SEPR first estimates the link forwarding probability based on history data. When two nodes meet, they exchange the link probability update messages called effective path length (EPL). A smaller EPL value suggests a higher probability of delivery. When a node received a smaller EPL, it will update its local EPL value. EPL is also used in deciding which nodes to forward the messages. Using SPER protocol, the same message could be forwarded to multiple nodes to increase reliability and reduce delay. Control-Movement Based In contrast to letting the mobile host wait passively for reconnection, the mobile hosts actively modify their trajectories to minimize transmission delay of messages. Some works have proposed approaches that try to limit delay by controlling node mobility. Zhao et al. (2004) propose a Message Ferrying (MF) approach for data delivery in sparse network. MF is a mobility-assisted approach which utilized message ferries to provide communication service for nodes in the network. Similar to their real life, message ferries move around the deployment area and take responsibility for carrying data between nodes. The main idea behind the Message Ferrying approach is to introduce non-randomness in the movement of nodes and exploit such non-randomness to help deliver data. Two variations of the MF schemes were developed. In the Node-Initiated MF (NIMF) scheme, ferries move around the deployed area according to known specific routes and communicate with other nodes when they meet. With knowledge of ferry routes, nodes periodically move close to a ferry and communicate with the ferry. In the Ferry-Initiated MF (FIMF) scheme, ferries move proactively to meet nodes. It is assumed that the ferry moves faster than nodes. When a node wants to send packets to other nodes or receive packets, it generates a service request and transmits it to a chosen ferry using a long range radio. Upon reception of a service request, the ferry will adjust its trajectory to meet up with the node and exchange packets using short range radios. In both schemes, nodes can communicate with distant nodes that are out of range by using ferries as relays. Zhao et al. (2005) also propose multiple ferries with stationary nodes to deliver data in networks and design of ferry routes. The route design problem with multiple ferries is more complicated than the single ferry case considering the possibility of interaction between ferries. The authors present four algorithms to generate ferry routes that meet the traffic demand and minimize the weighted delay. The authors considered algorithms that assume no interaction between ferries, either using a single route (SIRA) or multiple routes (MURA). In the single ferry, they adapt solutions for the wellstudied traveling salesman problem (TSP). In the multiple ferries, the algorithm is to assign nodes to specific ferries. The authors also considered algorithms that allow data relaying between ferries directly, ferry relaying algorithm (FRA), or indirectly, node relaying algorithm (NRA). In NRA, data is relayed between ferries via nodes, so the NRA adopts a geographic approach for assigning nodes to ferries. In FRA, data may be forwarded through multiple ferry routes while being routed to the destination. But instead of relaying data via 259 DTN Technologies for Vehicular Networks nodes as in NRA, ferries exchange data directly. Therefore, ferry routes need to be synchronized for two ferries to meet each other. Simulation results are obtained to evaluate the performance of route assignment algorithms, especially on the effect of the number of ferries on the average message delay. Numerical results indicate that when the traffic load is low, the improvement in delay due to the increased number of ferries is modest. This is because the delay is dominated by the distance between nodes. However, when the traffic load is high, an increase in the number of ferries can significantly reduce the delay. Flooding-based approach In the flooding-based approach, each node broadcasts the received packet to all of its neighbors, with the hope that one of these intermediate nodes will reach the destination. Epidemic Routing Epidemic routing is first proposed by Vahdat and Becker (2000) for forwarding data in a DTN. Epidemic routing utilizes the epidemic algorithm (Demers et al., 1987) that was originally proposed for synchronizing replicated databases. The epidemic algorithm ensures that a sufficient number of random exchanges of data in the network and guarantees all nodes will eventually receive all messages. Epidemic Routing works as follows. When two nodes come into contact, each node will exchange the list of all message IDs that they have in their buffers, called the summary vector, to see if there are any messages that the other node has that it has not received. After such pair-wise exchange of messages, each node will get all the messages carried by the other node that it has not received. When this operation completes, the nodes have the same messages in their buffers. The Epidemic Routing is similar to the flooding routing because it tries to send each message to all nodes using the summary vector in the net- 260 work. For this reason, Epidemic Routing incurs significant demand on both bandwidth and buffer. To reduce such overhead, there are many related paper to make epidemic routing consume fewer resources (Spyropoulos et al., 2005 ; Lindgren et al., 2003 ; Tan et al., 2003). To bound the overhead of delivering a message, Spyropoulos et al. (2005) proposed a technique called Spray and Wait to control the level of flooding. In the spray phase, there are L numbers of copies that are initially spread over the network by the source node or other nodes to L distinct relays. In the wait phase, if the destination was not found during spray phase, each node that has a copy of message will perform direct transmission. Binary spray and wait is a variation of Spray and Wait and produces a better performance. In this approach, the binary spray source node sends half of the copies of the message to the new relay node, and keeps the rest for itself. The source node and relay nodes repeat this procedure until there is only one copy left. When there is only one copy left, it switches to direct transmission. Conditional Routing In conditional routing, nodes are not blindly forward the messages to all or some neighbors. Instead, nodes estimate the probability of each link to destination and use this information to decide whether it should store the packet and wait for a better chance as well as to decide which nodes to forward. Lindgren et al. (2003) proposed a probabilistic routing protocol, called PROPHET (Probabilistic Routing Protocol using History of Encounters and Transitivity). PROPHET estimates probabilistic metric called delivery predictability. This metric indicates the probability of successfully delivering a message to the destination from the local node. PROPHET operates in a similar way as Epidemic Routing (Vahdat et al., 2000). When two nodes meet, they exchange summary vectors containing the delivery predictability vector which is based on the delivery predictability DTN Technologies for Vehicular Networks Table 1. The comparison of routing protocols for DTN Protocol Buffer Management Estimation of link forwarding probability Complexity of information exchange or computation for the link state Reactive or Proactive Epidemic Infinite No Don’t need Reactive CAR Infinite YES, using Kilman filter Computation only Reactive Spray and wait Infinite No Don’t need Reactive PROPHET Infinite YES, using delivery predictability vector Exchange and computation Reactive MaxProp Infinite YES, estimating the delivery likelihood Exchange and computation Reactive Knowledge Infinite YES, using oracle based Dijkstra algorithm Exchange and computation Reactive SEPR YES, remove those packets with smaller EPL YES computation Reactive Direction transmission Infinite Don’t need Don’t need Reactive MoVe Infinite Exchange and computation Exchange and computation Reactive NIFM YES, message time out and buffer overflow No Don’t need Proactive FIFN No No Compuation Proactive information. In theory, if two nodes are often encountered, they have high delivery predictability to each other. On the other hand, if a pair of nodes does not encounter each other in a while, they are intuitively not good forwarders of message to each other. Hence, the delivery predictability values must age (i.e. be reduced) as time goes. They showed in their simulation results that the communication overhead of PROPHET is lower than epidemic routing because PROPHET is only sent to better nodes. TransporT layer The existing transport layer protocols, such as TCP are not suitable for an environment where frequent disruption is a norm and end-to-end paths are typically not available. In (Ramadas et al., 2007), authors proposed the Licklider Transmission Protocol (LTP) that provides retransmission-based reliability over single links with extremely high latency such as deep space communications. LTP handles the transmission of a block of data, which can be split into segments that match the maximum transmission unit for the link. In an Interplanetary Internet setting, LTP is intended to serve as a reliable “convergence layer” protocol over single hop deep-space RF links. LTP implements ARQ of data transmissions by soliciting selective-acknowledgment reception reports. When a client service transmits the segments of data, some are flagged as checkpoints. When a checkpoint is received, the receiver returns a report of cumulative reception for that block. Reports acknowledge checkpoints and either signal successful reception or else trigger retransmission. Farrell et al. (2005) proposed a generic transport protocol for DTN that use the LTP extension mechanism (Farrell et al., 2007) to create an end-to-end capable transport protocol called LTP transport (LTP-T). The LTP extension mechanism was originally defined to handle the addition of authentication fields to LTP and al- 261 DTN Technologies for Vehicular Networks lows for the addition of both header and trailer extensions, up to a maximum of 16 (of each). In this work, authors define a new set of extensions of LTP about the transport protocol, i.e., source address, destination address, estimated block size and congestion notification etc. Since Bundle Protocol (Scott, K.et al., 2007) requires the services of a “convergence layer adapter (CLA)” which is an interface between the common bundle protocol and a specific internetwork protocol suite to send and receive bundles using an underlying Internet protocol, then in (Demmer et al., 2006) the authors present one such convergence layer adapter that uses the well-known Transmission Control Protocol (TCP). The TCP-based convergence layer (TCPCL) is used to link two bundle nodes. The lifetime of a TCPCL connection will match the lifetime of its underlying TCP connection. In other words, a TCPCL connection is initiated when a bundle node initiates a TCP connection to be established for the purposes of bundle communication. It is terminated either when the TCP connection ends due to one or both nodes actively terminating the TCP connection, or when network errors causes a failure of the TCP connection. In (Wood et al., 2007), the authors showed that the TCP protocol does not make effective use of available link capacity in a challenged environment like an opportunistic network. In (Wood et al., 2007) the authors proposed use Saratoga (Wood et al., 2008) as convergence layer. Saratoga is a UDP based file transfer protocol that can be used to transfer bundles. To send a bundle, the local bundle agent will either place bundles as files for Saratoga to transfer from its directory that can be accessible to both the bundle and Saratoga processes, or otherwise use inter-process communication to notify Saratoga of and provide a bundle to be transferred. 262 bundle layer As mentioned at section II, the bundle layer is responsible for storing, carrying and forwarding the data in DTN. Except from the unicast bundle delivery, multicast and anycast delivery approaches are typically used when there are more than one destination. bundle delivery approach In a DTN, applications utilize nodes to send or receive data that is carried in bundles which can be delivered to a group of nodes. When the group size is greater than one, the delivery semantics may be either the anycast or multicast. For anycast delivery, a bundle is delivered to at least one and preferable only one of the members in a group. On the other hand, for multicast delivery, the bundle is intended to be delivered to all members in the same multicast group. Anycast In (Gong et al., 2006), authors defined an anycast semantics model and propose a routing metric, called EMDDA (Expected Multi-Destination Delay for Anycast), for anycast. The semantics models allow message senders to explicitly specify the intended receivers of a message. In this study, the anycast routing algorithm is based on the metric EMMA which accurately estimates the delay from a node to the nearest member of the destined anycast group. EMDDA of a node to an anycast group is defined as the minimum value of Practical Expected Delays (PEDs), which PED is the expected delay of taking different paths, from the node to all the destination group members. When a message arrives at a node, but the node is not an intended receiver of the anycast message, the node will calculate its EMDDA to the destination group. DTN Technologies for Vehicular Networks Multicast Due to the network partitions and opportunistic contacts, nodes are difficult to maintain a sourcerooted multicast tree during the lifetime of a multicast session. The traditional approaches may fail to deliver a message when the link is unstable. In (Zhao et al., 2005), the authors investigated four approaches which are adopted from multicasting in the Internet or MANET. (1) UBR (UnicastBased Routing) uses unicast transfer to achieve the multicast service. (2) In Broadcast-Based Routing (BBR), messages will be flooded throughout the network in order to reach the intended receivers. (3) In Tree-Based Routing (TBR), messages are forwarded along a tree in the DTN graph that is rooted at the source and reaches all receivers. Messages are duplicated only at nodes that have more than one outgoing path. (4) Group-Based Routing (GBR) uses the concept of forwarding group which is a set of nodes that are responsible for forwarding the message. Messages will be flooded within the forwarding group to increase the chance of delivery. Based on the above algorithms, the authors present some algorithms for DTN are as follows. (1) UBR (Unicast-Based Routing) uses unicast transfer to each of the estimated intended receivers by encapsulating the original multicast message. The source node also buffers the multicast message and sends out new unicast messages when being informed of new intended receivers (2) In STBR (Static TreeBased Routing), messages are forwarded along a tree in the graph that is rooted at the source and reaches all receivers. Due to that the route is static, the message needs to wait for the next opportunity to be forwarded if a message misses a contact opportunity with a node. This may cause significantly increase the message delay. (3) In DTBR (Dynamic TreeBased Routing), nodes can determine the next-hop forwards of a message dynamically based on current available information. (4) BBR (BroadcastBased Routing) always includes all nodes in the network, so messages are flooded throughout the network. (5) GBR (GroupBased Routing) uses the concept of forwarding group for each message by computing a shortest path tree as in STBR. The group is a set of nodes that are responsible for forwarding the message. Messages are forwarded by flooding within the forwarding group. In (Ye et al., 2006), the authors proposed an on-demand situation-aware multicast (OS-mul- Table 2. The comparison of multicast protocols for DTN Protocol Knowledge Additional technologies Routing Type Tree Type UBR Unicast N/A N/A STBR Tree, using shortest path Source-routed, static tree N/A DTBR Tree, using shortest path Source-routed, dynamic tree BBR Broadcast N/A GBR Tree, using shortest path Source-routed, static path to a group OS-multicast Tree, using shortest path Source-routed, intermediate node dynamic rebuild the tree rooted itself CAMR Tree, using the contact probability estimate Source-routed Estimate the neighbor knowledge High power transmission and message ferry EBMR Tree, using the highest delivery predictability Source-routed delivery predictability Directional antenna N/A Complete knowledge or summary of the link states N/A N/A N/A 263 DTN Technologies for Vehicular Networks ticast) approach. Initially, a source-rooted tree is constructed in the similar way as STBR. When a node receives a bundle, it will dynamically rebuild the tree rooted at itself to all the destinations based on the current network conditions. Their simulation results show that OS-multicast can achieve smaller delays and better message delivery ratios than DTBR due to that each node in DTBR only forwards bundles to downstream nodes to reach the receivers in its receiver list and OS-multicast always use all possible chances to forward the bundles to all the destinations. In (Yang et al., 2006), the authors proposed a context-aware multicast routing (CAMR) scheme where nodes are allowed to use high power transmissions when the node density (which is locally observed) drops below a certain threshold. Each node maintains contact probabilities using its 2-hop neighbor information. This allows each node to deliver traffic without invoking a route discovery process if all receivers are within its two-hop neighbor. In addition, nodes are allowed to act as message ferries when they discover they are in a very sparse neighborhood and then travel to closer to the next-hop for delivering bundles. The combined high-power route discovery process and message ferrying features allow CAMR to achieve much higher multicast delivery ratio than DTBR and OS-multicast schemes. In (Chuah et al., 2007), the authors build the multicast scheme on top of the PROPHET (Lindgren et al., 2003), so this scheme is called encounter-based multicast routing (EBMR) scheme. This scheme has several enhancements to improve the delivery performance. First, each node selects as many nodes as needed with the highest delivery predictability to each of the multicast receivers. If the next-hop can not be found, a node will cache the data until the timer expires. When the timer expires, the node simply selects a node with the highest delivery predictability to multicast receivers. The second enhancement allows nodes 264 in the boundary region to use directional antenna to find nodes in other regions if they cannot hear any such node using omnidirectional antenna. security A draft DTNRG document presents the bundle security protocol specification (Symington et al., 2007) and an additional draft document (Symington et al., 2007) explains the rationale for the design choices made in the specification. The specification describes three security blocks that can be added to bundles to provide different security services. The Bundle Authentication block (BAB) is used to provide authentication over a single hop by adding a message authentication code or a signature to the bundle. The Payload Security block (PSB) is used to provide end-toend authentication in a similar fashion and the Confidentiality block (CB) is used to encapsulate encrypted payload of a bundle. Different combinations of these three security headers can be used simultaneously. Seth et al. (2005) propose the use of hierarchical Identity Based Cryptography (IBC) (Boneh et al., 2003) to achieve end-to-end security. The authors observe that traditional PKIbased approach is not suited for disconnected networks, since in DTN there do not have online access to an arbitrary receiver’s public key or certificate. In HIBC, different regions have sub-regions which maintain their own PKGs. A user sent messages with one PKG to a user of another PKG. The messages are authenticated and protected using the trust relations between PKGs and standard techniques of HIBC. The identifier of a principal can be based on existing well-known identifiers like e-mail addresses. However, in (Asokan et al., 2007), the authors argue that HIBC is not necessary because cellular operators already have roaming agreements intended to enable crossdomain operation. DTN Technologies for Vehicular Networks applicaTion layer In a DTN, traditional applications fail to take advantage of the communication opportunities offered by those opportunistic contacts. Hence, even the application is delay-tolerant in nature, the overall application performance can still suffer significantly in a disconnection-prone environment. We will discuss the application in DTN from the common example, e-mail, to the more complex example, web service. The well-known paradigms on the Internet is E-mail because this application is delay-tolerant by large and e-mail users are used to wait for hours or days for a reply. However, given that the underlying transport protocol of e-mail (i.e. TCP) is not designed for a DTN, supporting e-mail in such an environment is still quite challenging. Scott et al. (2005) proposed the use of DTN SMTP proxies to hide the disruptions between end users in a challenged network. This proxy is responsible to help the client to perform its work and exchanges the corresponding information to a peer proxy. The peer proxy receives the information and sends it to the SMTP server. The drawback of this proxy-based approach for SMTP protocol is that the proxy has to execute the entire SMTP protocol forwarding the information via the inter-proxy protocol. In (Hyyrylainen et al., 2007), the authors describe an architecture to enable mail communication in a heterogeneous environment that combines traditional server-based mail delivery and opportunistic communications for different types of devices. In this architecture, mail messages are sent in bundles into the DTN and carried toward a DTN mail gateway (DTNMWG). The DTN-MWG is responsible to forward and receive the mail between the infrastructure network and the DTN network. The DTN-MWG and corresponding device could implement the Bundle Protocol to eliminate unnecessary process (Scott et al., 2005). In addition, three different user equipment options can be configured on their device. (1) To enable the user to continue using traditional email applications, the user laptop can have a specialized email proxy to receive email from traditional and DTN connectivity options. (2) User equipment may contain separate mail application for both traditional mail and DTN mail. (3) User equipment may contain only the DTN mail application. Table 3. The comparison of application service for DTN Protocol Application Additional entities K. Scott E-mail DTN SMTP proxies Don’t modified Don’t modified T. Hyyrylainen et al. E-mail DTN mail gateway Add DTN Mail Proxy in client device Don’t modified DTN mail gateway Add another mail application for DTN mail Don’t modified DTN mail gateway Only receive the mail from DTN mail gateway Don’t modified Don’t modified Don’t modified Bundle router (BR) Add the bundle protocol Add the bundle protocol BR and Proxy Add the bundle protocol Add the bundle protocol Don’t modified Don’t modified Add the user query database Don’t modified K. Scott Web J. Ott et al. Web DTN-enable web proxy BR and Gateway A. Balasubramanian et al. Web Proxy Client Application Server 265 DTN Technologies for Vehicular Networks Supporting e-mail in a DTN system is straightforward since that fits into the characteristics of the DTN very well. Adding support for Web in DTN is much more complicated, because highly interactive application protocols, such as HTTP, are not well suited for this kind of environments. In (Scott et al., 2005), Scott proposed an implement of DTN-enable web proxy by extending the World Wide Web Offline Explorer (WWWOFFLE, 2009). The authors split the WWWOFFLE proxy and added client and a server side. The client side links to the challenged network and uses DTN bundles to communicate with the server side. The server side has full connectivity to the Internet, so that when the server receives requests from clients, it can use HTTP to retrieve the requested web pages through the Internet. In (Ott et al., 2006), authors presented a protocol design and a system architecture for delay-tolerant access to web pages. This work uses the bundle protocol to transport the HTTP payloads in DTN network. Furthermore, several scenarios are proposed for retrieving the content. First scenario is End-to-end HTTP-over-DTN. This scenario requires both client and server implementing HTTP-over-DTN so that bundles can be sent directly to the respective server. Second scenario is Proxy-based HTTPover-DTN. Adding proxies into an HTTP-overDTN may support a mobile node in content aggregation from one or more origin servers. Finally, gatewaying HTTP-over-DTN proposes gateway entities that communicate with another gateway through HTTP-over-DTN. The web clients and web servers communicate with gateway through HTTP-over-TCP and the intermediary gateway is required to covert between HTTP-over-DTN and HTTP-over-TCP operation. Balasubramanian et al. (2007) proposed a system, called Thedu which use an Internet proxy to collect search engine result and prefetch result pages. The mobile node can receive the user query through web interface and store it until the mobile node contacts with the proxy. If the connection is broken, the remaining web pages will be downloaded at next contact time. 266 Furthermore, when the proxy awaits connection from a mobile node and has pending response, it downloads the responses and fetches some relevant web pages. In addition, the proxy will prioritize response bundles at next contact time. conclusion DTN is an emerging system that is getting growing interest in networking research community. The DTN places different research challenges on different layers of a protocol stack. In this chapter, we provide a quick overview of the state-of-theart work in providing solutions to various issues in an opportunistic network. reFerences Asokan, N., Kostiainen, K., Ginzboorg, P., Ott, J., & Luo, C. (2007). Applicability of identity-based cryptography for disruption-tolerant networking. In . Proceedings of MobiOpp, 07, 52–56. doi:10.1145/1247694.1247705 Balasubramanian, A., Zhou, Y., Croft, B. W., Levine, B. N., & Venkataramani, A. (2007). Web search from a bus. In Proceedings of the second workshop on Challenged networks CHANTS CHANTS ’07, (pp. 59–66). Banerjee, N., Corner, M. D., & Levine, B. N. (2007). An energy-efficient architecture for dtn throwboxes. In . Proceedings of IEEE INFOCOM, 2007, 776–784. Boneh, D., & Franklin, M. (2003). Identity based encryption from the weil pairing. In SIAM Journal of Computing, 586–615. Brewer, E., Demmer, M., Du, B., Ho, M., Kam, M., & Nedevschi, S. (2005). The case for technology in developing regions. IEEE Computer, 38, 25–38. DTN Technologies for Vehicular Networks Burgess, J., Gallagher, B., Jensen, D., & Levine, B. N. (2006). Maxprop: Routing for vehicle-based disruption-tolerant networks. In . Proceedings of IEEE INFOCOM, 2006, 1–11. doi:10.1109/ INFOCOM.2006.228 Burns, B., Brock, O., & Levine, B. N. (2005). Mv routing and capacity building in disruption tolerant networks. Proceedings of IEEE INFOCOM, 1, 398–408. Chaintreau, A., Hui, P., Crowcroft, J., Diot, C., Gass, R., & Scott, J. (2005). Pocket switched networks: Real-world mobility and its consequences for opportunistic forwarding. Technical report, University of Cambridge Computer Laboratory, Cambridge, UK. Chen, Y., Zhao, W., Ammar, M., & Zegura, E. (2007). Hybrid routing in clustered dtns with message ferrying. In Proceedings of the 1st international MobiSys workshop on Mobile opportunistic networking MobiOpp ’07, (pp. 75–82). Chuah, M., & Xi, Y. (2007). An encounter-based multicast scheme for disruption tolerant networks. Technical report published by the University of Lehigh, Lehigh, PA. Demers, A. A., Greene, D., Hauser, C., Irish, W., Larson, J., Shenker, S., et al. (1987). Epidemic algorithms for replicated database maintenance. In Proceedings of the Sixth Symposium on Principles of Distributed Computing, (pp. 1–12). Demmer, M., & Ott, J. (2006). Delay tolerant networking tcp convergence layer protocol. Internet-draft. Fall, K. (2003). A delay-tolerant network architecture for challenged internets. In Proceedings of ACM SIGCOMM, (pp. 27–34). Farrell, S., & Cahill, V. (2005). Ltp-t: A generic delay tolerant transport protocol. Technical report, University of Dublin, Dublin. Farrell, S., Ramadas, M., & Burleigh, S. (2007). Licklider transmission protocol - extensions. Internet-draft. Freitag, L., Grund, M., Singh, S., Partan, J., Koski, P., & Ball, K. (2005). The whoi micro-modem: An acoustic communications and navigation system for multiple platforms. In Proceedings of the IEEE/MTS OCEANS Conference, Washington DC, (Vol. 2, pp. 1086–1092). Gong, Y., Xiong, Y., Zhang, Q., Zhang, Z., Wang, W., & Xu, Z. (2006). Anycast routing in delay tolerant networks. In IEEE GLOBECOM (pp. 1–5). Grossglauser, M., & Tse, D. N. C. (2002). Mobility increases the capacity of ad-hoc wireless networks. In IEEE/ACM Transactions on Networking, 10(4), 477–486. Hyyrylainen, T., Karkkainen, T., & Luo, C. (2007). Opportunistic email distribution and access in challenged heterogeneous environments. In Proceedings of the second workshop on Challenged networks CHANTS (pp.97–100). Interplanetary internet project (n.d.). Interplanetary internet project. Retrieved from http://www. ipnsig.org Ip, Y. K., Lau, W. C., & Yue, O. C. (2007). Forwarding and replication strategies for dtn with resource constraints. In . Proceedings of IEEE Vehicular Technology Conference, 1, 1260–1264. Jain, S., Fall, K., & Patra, R. (2004). Routing in a delay tolerant network. In Proceedings of ACM SIGCOMM, (Vol. 34, pp. 145–158). Jun, H., Ammar, M. H., Corner, M. D., & Zegura, E. W. (2006). Hierarchical power management in disruption tolerant networks with traffic-aware optimization. In Proc. ACM SIGCOMM Workshop on Challenged Networks (CHANTS), (pp. 245–252). 267 DTN Technologies for Vehicular Networks Jun, H., Ammar, M. H., & Zegura, E. W. (2005). Power management in delay tolerant networks: A framework and knowledge-based mechanisms. In IEEE Communications Society Conference on Sensor and Ad Hoc Communications and Networks (SECON), (pp. 418–429, 2005). Krotkov, E., & Blitch, J. (1999). The defense advanced research projects agency (darpa) tactical mobile robotics program. In The International Journal of Robotics Research (pp. 769–776). LeBrun, J., Chuah, C.-N., Ghosal, D., & Zhang, M. (2005). Knowledgebased opportunistic forwarding in vehicular wireless ad hoc networks. In IEEE Vehicular Technology Conference (VTC), (pp. 2289– 2293). Leguay, J., Friedman, T., & Conan, V. (2006). Evaluating mobility pattern space routing for DTNs. In Proceedings of IEEE INFOCOM. Lindgren, A., & Doria, A. (2007). Experiences from deploying a reallife dtn system. In 2007 4th IEEE Consumer Communications and Networking Conference, (pp. 217–221). Lindgren, A., Doria, A., & Schelen, O. (2003). Probabilistic routing in intermittently connected networks. ACM SIGMOBILE Mobile Computing and Communications Review, 7(3), 19–20. doi:10.1145/961268.961272 Musolesi, M., Hailes, S., & Mascolo, C. (2005). Adaptive routing for intermittently connected mobile ad hoc networks. In IEEE WoWMoM 2005 (pp. 183–189). Ott, J., & Kutscher, D. (2006). Bundling the web: Http over dtn. In Proceeding of WNEPT2006. Ramadas, M., Burleigh, S., & Farrell, S. (2007). Licklider transmission protocol - specification. Internet-draft. 268 Rodoplu, V., & Park, M. K. (2005). An energyefficient mac protocol for underwater wireless acoustic networks. In Proceedings of the MTS/ IEEE OCEANS’05 Conference. Scott, K. (2005). Disruption tolerant networking proxies for on-the-move tactical networks. In IEEE MILCOM 2005, (Vol. 5, pp. 3226–3231). Scott, K., & Burleigh, S. (2007). Bundle protocol specification. Internet- Draft. Segui, J., & Jennings, E. (2006). Delay tolerant networking v bundle protocol simulation. In Second IEEE International Conference on Space Mission Challenges for Information Technology. Seth, A., & Keshav, S. (2005). Practical security for disconnected nodes. In Proceeding of Workshop on Secure Network Protocols (pp. 31–36). Spyropoulos, T., Psounis, K., & Raghavendra, C. S. (2004). Single-copy routing in intermittently connected mobile networks. In 2004 First Annual IEEE Communications Society Conference on Sensor and Ad Hoc Communications and Networks, (pp. 235 – 244). Spyropoulos, T., Psounis, K., & Raghavendra, C. S. (2005). Spray and wait: An efficient routing scheme for intermittently connected mobile networks. In Proceeding of the 2005 ACM SIGCOMM workshop on Delay-tolerant networking WDTN ’05, (pp. 252–259). Symington, S., Farrell, S., Weiss, H., & Lovell, P. (2007). Bundle security protocol specification. Internet-draft. Symington, S., Farrell, S., Weiss, H., & Lovell, P. (2007). Delay-tolerant networking security overview. Internet-draft. Tan, K., Zhang, Q., & Zhu, W. (2003). Shortest path routing in partially connected ad hoc networks. In Proceedings of IEEE Globecom 2003, (Vol. 2, pp. 1038–1042). DTN Technologies for Vehicular Networks Tariq, M. M. B., Ammar, M., & Zegura, E. (2006). Message ferry route design for sparse ad hoc networks with mobile nodes. In Proceedings of the seventh ACM international symposium on Mobile ad hoc networking and computing MobiHoc ’06 (pp. 37–48). The zebranet wildlife tracker. (2009). Retrieved from http://www.princeton.edu/ mrm/zebranet. html Tier. (2009). Retrieved from http://tier.cs.berkeley. edu/wiki/Home Vahdat, V., & Becker, D. (2000). Epidemic routing for partially connected ad hoc networks. Technical Report CS-200006, Duke University. Wills, J., Ye, W., & Heidemann, J. (2006). Lowpower acoustic modem for dense underwater sensor networks. In Proceedings of the First ACM International Workshop on UnderWater Networks (WUWNet), Los Angeles, CA, (pp. 79–85). New York: ACM. Retrieved from http://www.isi.edu/ ohnh/PAPERS/Wills06a.html Wood, L., Eddy, W. M., Ivancic, W., McKim, J., & Jackson, C. (2007). Saratoga: a delay-tolerant networking convergence layer with efficient link utilization. In Proceeding of IWSSC’07. Wood, L., McKim, J., Eddy, W., Ivancic, W., & Jackson, C. (2008). Saratoga: a convergence layer for delay-tolerant networking. Internet draft. WWWOFFLE. (2009). Retrieved from http:// www.gedanken.demon.co.uk/wwwoffle/ Yang, P., & Chuah, M. C. (2006). Context-aware multicast routing scheme for disruption tolerant networks. In Proceedings of ACM international workshop on PE-WASUN ’06 (pp. 66–73). Ye, Q., Cheng, L., Chuah, M. C., & Davison, B. D. (2006). Os-multicast: On-demand situation-aware multicasting in disruption tolerant networks. In . Proceedings of IEEE VTC, 1, 96–100. Zhang, Z., & Fei, Z. (2007). Route design for multiple ferries in delay tolerant networks. In Proceedings of IEEE Wireless Communications and Networking Conference (pp. 3460–3465). Zhao, W., Ammar, M., & Zegura, E. (2004). A message ferrying approach for data delivery in sparse mobile ad hoc networks. In Proceedings of the 5th ACM international symposium on Mobile ad hoc networking and computing MobiHoc ’04 (pp. 187–198). Zhao, W., Ammar, M., & Zegura, E. (2005). Controlling the mobility of multiple data transport ferries in delay-tolerant network. In . Proceedings of IEEE INFOCOM, 2, 1407–1418. Zhao, W., Ammar, M., & Zegura, E. (2005). Multicasting in delay tolerant networks: Semantic models and routing algorithms. In Proceeding of Sigcomm Workshop in DTN, (pp. 268–275). Wood, L., Peoples, C., Parr, G., Scotney, B., & Moore, A. (2007). Tcps protocol radius:the distance where timers prevent communication. In Proceeding of IWSSC’07. 269 Section 6 Management and Traffic Control 271 Chapter 16 Simple Transportation Management Framework Chyi-Ren Dow Feng Chia University, Taiwan, R.O.C. absTracT The Simple Transportation Management Framework (STMF) specifies a set of rules and protocols which can be used to organize, describe, and exchange transportation management information between transportation management applications and equipments. The STMF framework consists of four elements, including Management Information Base (MIB), Structure and Identification of Management Information (SMI), Simple Network Management Protocol (SNMP), and Simple Transportation Management Protocol (STMP). MIB is a collection of management objects written in ASN.1 notation. SMI is the definition of how to create management objects and a hierarchical definition of nodes where management objects will be attached for unique identification. SNMP is a communications protocol for configuring and monitoring of network devices. STMP is a variation of SNMP to address low-bandwidth communication links and real-time device monitoring. inTroducTion The STMF specifies a set of rules and protocols which can be used to organize, describe and exchange transportation management information between transportation management applications and equipments. STMP is based on the Internetstandard Network Management Framework and its purpose is to provide a high-level compatibility, DOI: 10.4018/978-1-60566-840-6.ch016 inter-operability, and maintenance of STMF. The STMF framework consists of the following four elements. The brief definitions of these elements are listed below and the detailed descriptions and relationships of these elements will be given in the following sections. • Management information base (MIB) (Perkins et al., 1997): A collection of management objects written in ASN.1 (Abstract Syntax Notation One) (Steedman et al., Copyright © 2010, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global is prohibited. Simple Transportation Management Framework Figure 1. Relationships of SMI, MIB, and ASN.1 • • • 1993; ITU-T X.680-X.690, 1994) notation, which is a standard and flexible notation that describes data structures for representing, encoding, transmitting, and decoding data. Structure and identification of management information (SMI): The definition of how to create management objects and a hierarchical definition of nodes where management objects will be attached for unique identification. Simple network management protocol (SNMP) (Stallings et al., 1993; Stallings et al., 1996; Feit et al., 1995): A communications protocol developed by the IETF for configuring and monitoring of network devices. Simple transportation management protocol (STMP): A variation of SNMP developed by NEMA to address low-bandwidth communication links and real-time device monitoring. NEMA is the trade association of choice for the electrical manufacturing industry, and it provides a forum for the development of technical standards relationships of smi, mib, and asn.1 SMI describes the common structures and identification schemes for the definition of management information. ASN.1 is used to specify SMI and it can be compiled by MIB compilers. To do so, SMI 272 definitions (ASN.1 specifications) are included in a MIB module. As shown in Figure 1, the difference is made because SMI defines how to create managed objects and how to utilize ASN.1 in order to create and identify management information (MIB objects) within a tree-like structure. Management center uses Structure and Identification of Management Information to define MIB, then using BER or OER encoding scheme to generate SNMP or STMP. smi Managed objects would be accessed via MIB and objects in the MIB would be defined using ASN.1 which should be in conformance with IAB STD 16 (RFC 1212). Each object type would have a name, syntax, and an encoding. The OBJECT IDENTIFIER would represent a unique name. An OBJECT IDENTIFIER should be administratively assigned a name. The administrative policies discussed in RFC 1212 would be used for assigning names and identifiers. When transmitted on the network, the encoding of an object type determines how its instances are represented. Names Names are used to identify managed objects. This sub-clause specifies names that should be hierarchical in nature. The OBJECT IDENTIFIER concept is used to model this notion. OBJECT IDENTIFIERs can be used to identify objects, Simple Transportation Management Framework Figure 2. nema of OBJECT IDENTIFIER In this example, four nodes are assigned under the nema node. The definitions are listed below. • • • • nemaMgmt OBJECT IDENTIFIER::= {nema 1} nemaExperimental OBJECT IDENTIFIER::= {nema 2} nemaPrivate OBJECT IDENTIFIER::= {nema 3} transportation OBJECT IDENTIFIER::= {nema 4} SYNTAX regardless of the semantics associated with the object (e.g., a network object, a standards document, etc.). An OBJECT IDENTIFIER is a sequence of integers which can be used to traverse a global tree. The tree consists of a root connected to a number of labeled nodes via edge and each node may have branches of its own that are labeled. This process may continue to an arbitrary level of depth. A label is a pairing of a brief textual description and an integer. The format used to show this pairing shall be text (integer). The IAB has transferred authority for the nema node to NEMA. As shown in Figure 2, the nema node is defined below using the ASN.1 syntax. nema OBJECT IDENTIFIER::= { iso(1) org(3) dod(6) internet(1) private(4) enterprises(1) 1206 } The SYNTAX is used to define the structure and it is corresponding to object types. RFC 1155 can be used to define this structure. The SYNTAX type consists of primitive types and guidelines for enumerated INTEGERs and constructor types.16.1.2.3 Managed Objects IAB STD 16 specifies a format to be used by documents that defines objects. All objects would be defined using the OBJECT-TYPE macro described in RFC 1212. The OBJECT-TYPE macro definition contains at least five fields, including object name, syntax, access, status and description. Figure 3 shows an example of managed object. The “Access” can be read-only, read-write, write-only or no-accessible. The “Status” can be mandatory, optional, obsolete, or deprecated. mib The MIB version 2, published in RFC 1213 in 1991, steped up a number of useful variables missing from MIB-I. It has four characteristics: the first one is that has no relationship with network management protocol (ITS Standards et al., 2003; Aidarous et al., 1994). The second is each object has a unique Object Identifier (OID). The third is that defined by SMI. The final one records the type and access limits of authority. 273 Simple Transportation Management Framework Figure 3. ASN.1 Macro As shown in Figure 4, the Object Naming Tree is used to record all objects in MIB-II. If we choose tcpConnState for instance, the location of tcpConnState is iso(1)→ org(3)→ dod(6)→ internet(1)→ mang(2)→ mib-2(1)→ tcp(6)→ tcpConnTable(4)→ tcpConnState(1), so we can know that the OID of tcpConnState is 1.3.6.2.1.6.4.1. From Figure 4, we can also see some groups in MIB-II, including system(1), interface(2), at(3), ip(4), icmp(5), tcp(6), udp(7), egp(8), cmot(9) and transmission(10). Figure 4. Object naming tree 274 Transportation management protocol Figure 5 shows the comparison of SNMP, STMP and SFMP. The STMP is based on SNMP. Several features are modified so that STMP can provide more efficient operations. STMP can be used to save bandwidth and offer all functions of SNMP under the infrequent information demands. Thus, STMP is the subset of SNMP and, the STMP managed system can also use SNMP to carry on the communications. The major advantage of STMP is that it uses an effective code method that is sup- Simple Transportation Management Framework Figure 5. Comparisons of SNMP, STMP and SFMP ported by dynamic objects to reduce the overhead of package. The user can also set any information object by herself/himself through dynamic objects. Thus, STMP is the choice which is more elasticity and saves more bandwidth. However, compared with SNMP, it is much more difficult in practice for STMP. TMP allows three protocols to coexist while using the same protocol identifier. This was achieved as a result of the fact that all SNMP messages start with an initial byte of 0x30 (e.g., SNMP uses Basic Encoding Rules and all SNMP messages are defined as a SEQUENCE of data). TMP has been designed to use the value of this first byte to identify which protocol is being referenced. 0x30 is a value which identifies an SNMP message. Both SFMP and STMP messages use the highest order four bits of the first byte to determine the type of message (e.g., get request and set request). The lowest order four bits are used to identify whether the message is a fixed message or one of the 13 dynamic objects. The table in Figure 6 defines the specific mapping of the first byte value. snmp IETF defined SNMP which is a part of the protocol suite to describe the managed objects contained in MIB. This simple protocol is used in network management systems to monitor network-attached devices for conditions that warrant administrative attention. It consists of a set of standards for network management such as application layer protocol, a database schema, and a set of MIB. snmp architectural model The SNMP architectural model is a collection of network management stations and network elements. It consists of three key elements, including Management stations, multiple network elements, and Network Management Protocol. A management station can perform management applications which monitor and control network elements. Multiple network elements are devices such as hosts, routers, and gateways that contain agents to perform the network management functions requested by the network management stations. The last, Simple Network Management Protocol is used to exchange network management information between the network management stations and the agents. The SNMP architectural model is shown in Figure 7. The upper parts of the management station and network elements are manager and agents, respectively. The lower part is responsible for the information exchange between management station and the different network elements. Each component of SNMP is described in the next section. 275 Simple Transportation Management Framework Figure 6. Mapping of the first byte value management components There are four management components in SNMP, including Manager, Agent, Network Protocol and Management Information Base. The components are described as follows: 276 • • Manager: Manager station that runs management applications to provide an interface. The network manager can monitor and control the network by the interface. Agent: The software that responds to request for information from the manager station and according to the request from manager station to act. Therefore, network Simple Transportation Management Framework Figure 7. SNMP Architectural Model Figure 8. MIB architectural diagram • • device should be equipped with agent software so that they can be managed from a manager station. Network management protocol: The communication protocol between manager stations and agents. It provides a standard way to exchange information between manager stations and agents. Management information base: A collection of Manager Objects (MO) that are managed in remote devices. As shown in Figure 8, each MO is a data variable and each agent is a network that maintains an MIB. A manager action is either “monitor” or “control”. To monitor resources in a remote device, we can read the value of MOs in the MIB. In the MIB, we can modify the values of MOs to “control” resources in a remote device. snmp service SNMP provides four services and five SNMP Protocol Data Units (PDUs). The four services are Get, Set, GetNext and Trap. The five SNMP PDUs are GetRequest, SetRequest, GetNextRe- quest, GetResponse, and Trap. Figure 9 shows the relationship between services and management elements. Manager sends commands such as Get, Set, GetRequest, SetRequest and GetNextRequest to agents and then agent responds GetResponse. If agent occurs error during runtime, agent application submits Trap to manager. There are two versions of SNMP, including SNMP v1 and SNMP v2. SNMP v2 provides more services than SNMP v1. SNMP v1 As shown in Figure 10, four services provided by SNMP v1 are Get, Set, GetNext and Trap. Below 277 Simple Transportation Management Framework Figure 9. Services of SNMP Figure 10. SNMP v1 services are detailed descriptions about the five SNMP PDUs: GetRequest, SetRequest, GetNextRequest, GetResponse, and Trap. • • • • • Get Request: Used to retrieve the values of objects in the MIB of an agent. Get-Next Request: Used to retrieve the values of the next objects in the MIB of an agent. The management station can use this command through all MIB tree paths to retrieve all variables. Get Response: When an agent got Get Request from manager, the agent responds the value to manager. Set Request: Used to set or modify the values in the MIB of an agent. Trap: Used to report extraordinary events to the manager if an error occurs on the remote device. Figure 11. SNMP packet 278 SNMP v2 There are two new services provided by SNMP v2: GetBulk and Inform. Centers use GetBulk command to retrieve large amount of MIB objects data while different devices use Inform command to provide information to the management center. snmp packet As shown in Figure 11, the SNMP packet consists of three fields: Version, Community Name and Protocol Data Unit. • • Version: The number of SNMP version that is used by Manger and Agent. The “0” is used to indicate SNMP v1. The “1” is used to indicate SNMP v2. Community name: It is the password of SNMP. The Manager and Agent should Simple Transportation Management Framework Figure 12. PDU for GetRequest, GetNextRequest and SetRequest • use the same Community Name or else the frame will be discarded. Protocol data unit (PDU): It consists of some data that would be used in communication such as Request ID, Error Status and VarBindList. PDU As shown in Figure 12, the PDU for Get and Set service has four fields, including Request ID, Error Status, Error Index and VarBindList and these fields are described in detail as follows: • • RequestID: An integer type is used as an identification. It identifies different service actions in communication for SNMP operation. Error Status: Used to confirm whether this PDU is correct. The value will be equal to zero if this is a “Get-Request” command. If this command is “Get-Response”, the value of Error status will have six possible values as shown in Table 1. Error Index indicates that the index of VarBindList is wrong. The type of Error Index is integer. For the request from management, the values of Error Status and Error Index will be 0. VarBindList is a list that combines Variable IDs and Variable Values. The architecture of VarBindList is shown in Figure 13. The variable ID field is an Object ID and the field of Variable Value may be an integer, octstring or IP Address. The Object ID and the type of Variable Value are defined in MIB as mentioned in MIB earlier in the chapter. SNMP Trap PDU The Trap command is submitted from agent. The architecture of Trap PDU is shown in Figure 14. The Trap PDU consists of six fields, including Enterprise, Agent Address, Generic Trap Number, Specific Trap Number, Time Stamp and VarBindList. The detailed descriptions of the six fields are as follows: • • Table 1. Values of error status and its meaning Value of Error Status Meaning 0 No Error 1 PDU has too many bytes 2 There is no object with this name 3 Identifying the PDU type is bad 4 Incorrect implementation of SNMP 5 Unspecified errors of other types • • • Enterprise: Type of object generating trap. Agent address: Address of object generating trap. Generic trap number: Using integer to represent seven Trap types defined in RFC 1157. Specific trap number: The specific Trap type defined by enterprise. Time stamp: Time elapsed between the last initialization of the network entity and the generation of the trap. 279 Simple Transportation Management Framework Figure 13. VarBindList Set-Request Command Set command writes MIB object value. For example, if we have to change the system description (sysDescr), the command can be set as below: • VarBindList: Specific or all information that can be used to solve this trap problem. Get-Request Command SNMP uses index object to select column. The index object of ipRouteTable is the ipRouteDestination and the entity of ipRouteNextHop is 1.3.6.1.2.1.4.21.1.7.10.3.4.5. While using SNMP command interface, the Get-Request command should be inserted as below: • • Get-Request(1.3.6.1.2.1.4.21.1.7.10.3.4.5) or Get-Request(ipRouteNextHop.10.3.4.5) Get-Next-Request Command GetRequest command is used to request one object of a device per time. Thus, to request all of the devices, Get-Next_Request can be used. The command can be inserted as below: • • Get-Next-Request (1.3.6.1.2.1.4.21.1.7.10.3.4.5)or Get-Next-Request (ipRouteNextHop.10.3.4.5) Figure 14. SNMP trap PDU 280 Set-Request (sysDescr = name) Set-Request command in SNMP can be used to add or delete rows of table. The result of the SetRequest is related to the real implementation. SNMP Example Figure 15 shows a SNMP example of a traffic light controller. The center gets the remaining time from the traffic light controller. MIB can only send a request each time. While the traffic center requests information from the controller, the traffic controller will request a DeviceNum from the center. Then, the center will send TimePhaseNum to the controller and wait for response. The traffic center can request MaxGreenTime, MinGreenTime, YellowTime, GreenTime, RedTime from the traffic light controller. Each request will get a response from the controller in the same way. sTmp This section introduces the Simple Transportation Management Protocol and the comparison between Simple Transportation Management Protocol and Simple Network Management Protocol. STMP is based on SNMP and supports dynamic object that is a group of data elements. Simple Transportation Management Framework Figure 15 Example of a traffic light controller Besides, it also supports PMPP defined in the sub network of NTCIP. dynamic object Dynamic object is a simple grouping of data elements. It consists of a set of related MIB objects. For example, object name, object phase, and object time value are dynamic objects of the time controller of a traffic light. These three objects are related to each other. Between management station and the agent, dynamic object increases control flexibility. A dynamic object can combine all related objects into a set. Thus, different tasks can be done easier using a dynamic object. To reduce the need of bandwidth during the transformation, dynamic object is defined at run time. It is an effective way for the management station to communicate with different agents. However, dynamic object can only access using the STMP protocol. Dynamic Object Configuration Table Dynamic object configuration table consists of three elements: dynObjNumber, dynObjConfigOwner, and dynObjConfigStatus. The dynObjNumber provides a number for the frequently used dynamic objects. The main purpose of the dynObjNumber is to identify which of the 13 dynamic objects this row of the table is associated. Each dynamic object has an owner. Therefore, dynObjConfigOwner is used to indicate the identity of the owner that defined the dynamic object. Dynamic object records status of each managed objects. Thus, dynObjStatus is used to indicate the status of the dynamic object. Table 2 depicts the composite table for dynamic object configuration and definition. The table includes the three dynamic object (dynObjNumber, dynObjConfigOwner, and dynObjConfigStatus), dynObjIndex, and dynObjVariable. 281 Simple Transportation Management Framework Table 2. Composition table for dynamic object dynObjNumber dynObjConfigOwner 1 <Owner of Dynamic Object #1> dynObjConfigStatus <Status of Dynamic Object #1> 2 <Owner of Dynamic Object #2> <Status of Dynamic Object #2> 3 <Owner of Dynamic Object #3> <Status of Dynamic Object #3> … … 13 <Owner of Dynamic Object #13> … <Status of Dynamic Object #13> The dynamic object is the combination of managed objects. Each dynamic object is given an index to represent the sequence of the objects. The dynObjIndex column indicates the index of the dynamic objects. The dynamic object method treats all frequently used related objects as a set of dynamic objects. 282 dynObjIndex dynObjVariable 1 <OID of 1st object in dynObj 1> 2 <OID of 2nd object in dynObj 1> 3 <OID of 3rd object in dynObj 1> … … 255 <OID of 255th object in dynObj 1> 1 <OID of 1st object in dynObj 2> 2 <OID of 2nd object in dynObj 2> 3 <OID of 3rd object in dynObj 2> … … 255 <OID of 255th object in dynObj 2> 1 <OID of 1st object in dynObj 3> 2 <OID of 2nd object in dynObj 3> 3 <OID of 3rd object in dynObj 3> … … 255 <OID of 255th object in dynObj 3> … … 1 <OID of 1st object in dynObj 13> 2 <OID of 2nd object in dynObj 13> 3 <OID of 3rd object in dynObj 13> … … 255 <OID of 255th obj dynObj 13> Objects in the set have already defined in the ISO naming tree. Thus, dynObjVariable records each managed object’s OID, which is the location of the managed object in the ISO naming tree. Simple Transportation Management Framework Dynamic Objects and System Operation STMP supports 13 dynamic objects for each agent. With a different set of dynamic objects, the management station could configure each device. In practice, most management stations are likely to configure similar devices with similar dynamic object definitions. State Transition for dynObjConfigStatus Table 3 shows the state transition for dynObjConfigStatus. If no action takes place and response indicates no Error. When the state changes to invalid, all entries associated with the ConfigEntryStatus object are deleted or cleared and response indicates no error. If no action takes place but response indicates badValue. If Dynamic Object Validation succeeds then state changes to valid and response indicates no error. If Dynamic Object Validation fails then state remains underCreation and response indicates genErr. The state changes to underCreation and the response indicates no error. Set Request-No Reply is similar to Set Request, but the device will not return any response after the request is sent. Get Response and Set Response are used to response to get and set requests command from centers. Trap Response is to inform the centers that some situations occurred on a device. Get Error Response will response to the centers that get request command is failed. Set Error Response will response to the centers that set request command is failed. STMP PDU Field Figure 16 shows the STMP PDU fields. There are four fields in the STMP PDU: PDU format, message type, object ID and information field. The PDU Format is 0 or 1 (bit 7). If PDU format is “1”, it represents the STMP service. If the data is “0”, it represents a SNMP service or it might be reserved for future use. Message type and object ID are header fields. Message type is represented by bits 6-4. It indicates the STMP service. Below are functions of different message types: STMP Service • Services provided by STMP are Get Request, Set Request, and Set Request-No Reply. The service responses are Get Response, Set Response, Trap Response, Get Error Response, and Set Error Response. Get Request is to get an object’s information. Set Request is used to set an objects information. • • • 000: An STMP-GetRequest-PDU is contained in the packet. 001: An STMP-SetRequest-PDU is contained in the packet. 010: An STMP-SetRequest-NoReply-PDU is contained in the packet. 011: An STMP-GetNextRequest-PDU is contained in the packet. Table 3. State Transition for dynObjConfigStatus Requested State Invalid Current State Invalid UnderCreation Valid Invalid (1) UnderCreation (2) Invalid (3) UnderCreation Invalid (2) UnderCreation (3) Valid (2) or underCreation Valid Invalid (2) Valid (3) Valid (1) 283 Simple Transportation Management Framework Figure 16. PDU field • • • • 100: An STMP-GetResponse-PDU is contained in the packet. 101: An STMP-SetResponse-PDU is contained in the packet. 110: An STMP-ErrorResponse-PDU is contained in the packet. 111: Reserved by TMP for future use. The object ID carries the ID of the object. It is represented by bits 3-0. Below are descriptions about the data of object ID: • 0000: Reserved by TMP for SFMP. Figure 17.Configuring time dynamic object 284 • • • 0001-1101: ID of STMP 13 dynamic objects 1110: Reserved by TMP for future use. 1111: Reserved by TMP for future use. The information field contains information that has to send between the server and the agent. It would be empty for STMP-GetRequest, STMPGetNextRequest, and STMP-SetResponse. The PDU Information field for STMP-GetResponse, STMP-SetRequest, and STMPSetRequestNoReply would consist of a series of component fields, each encoding one Referenced Object. Besides, Simple Transportation Management Framework the component fields would be encoded in order. It is according to the associated dynObjIndex, with the first field encoding the value of the first Referenced Object of the Dynamic Object, and the last field encoding the value of the last Referenced Object of the dynamic object. Examples The examples in the following will provide more detailed explanation about the STMP and the dynamic objects. Example: Configuration Time Dynamic Object Before using STMP to transfer object, we should set Dynamic Object Configuration Table. For example, we configure the third object as Time object. As shown in Figure 17, we should set the third object is “invalid” first, and then we set it as “underCreation”. Now we just can set the property of dynamic object as the following: • • Set the owner of this dynamic object to “FCU”. Set the variables of the dynamic object with MIB object. In this example, the dynamic object includes “GlobalTime” object as variable 3.1 and “GlobalDayLightSaving” • object as variable 3.2 and so on. This can be shown in Figure 17. Finally, set this dynamic object to “valid”. Now we can use this dynamic object to transfer data. Example: Getting a Dynamic Object After finishing the setting of dynamic object, we can use this dynamic object to transfer data with services such as “request” service. To continue above example, we are going to discuss the details of STMP PDU with request command. As shown in Figure 18, we can see the actual content of STMP “GetRequest” and “GetResponse” PDU. In the Information fields, there are four variables that we set in above example. They are presented as the following: • • • • 3A 24 63 20 variable 1 = globalTime.0 = November 29, 2000 at 2:00 am UTC 03 variable 2 = globalDaylightSaving.0 = 3 = enableUSDST FF FF B9 B0 variable 3 = controller-standardTimeZone.0 = -18000 = EST 06 53 61 6D 70 6C 65 variable 4 = eventClassDescription.1 (6 bytes) = “Sample” Figure 18. Actual content of STMP PDU 285 Simple Transportation Management Framework Application Level: STMP Figure 19 shows an example of configuring traffic light controller of dynamic object. First, the traffic center sends a request to the traffic light controller. The request demands the traffic light controller to set the dynObjConfigStatus as “invalid”. The center sends another request to the controller to set the dynObjConfigStatus as “underCreation” when the traffic center gets response from the traffic light controller. The dynamic object can be set after the dynObjConfigStatus is set as “underCreation”. The traffic center starts setting the dynamic objects (e.g., dynObjConfigOwner and dynObjVariable) when the traffic light controller sends a response. After all dynamic objects are set; the traffic center has to set the dynObjConfigStatus as “valid”. Finally, setting the dynamic objects has done. After the setting of dynamic object is finished, we can use it to actual applications. An applica- Figure 19. Example of configuring traffic light controller of dynamic object Figure 20. Application for traffic light using dynamic object 286 Simple Transportation Management Framework Figure 21. Comparisons between SNMP and STMP tion for traffic light is shown in Figure 20. This application uses dynamic object to control traffic light. Figure 20 shows the traffic center submits a get request to the traffic light controller and then the traffic light controller returns the value of traffic light. In Figure 20, the “dynObjVariable” word is the object number of dynamic object. The traffic center uses getRequest service with object number to request the value of traffic light; the traffic light controller would then responses to the traffic center with “getResponse” command. As shown in Figure 21, the difference between SNMP and STMP package is that SNMP has Version number and Community Name as a password. Because SNMP uses more bandwidth than STMP. SNMP can be used in the environment that has weak probability of controlling object and high network bandwidth. In comparison with SNMP, STMP uses dynamic objects to reduce the data transfer and efficiently use bandwidth. Therefore, STMP can be used in the environment and it has low network bandwidth and frequent control of object value. In summary, SNMP is simple and easy to implement. STMP uses Dynamic Composite Objects so that STMP uses bandwidth efficiently. aTms system for sTmp The Advanced Traffic Management System (ATMS) (National Highway Institute, 2001) is based on an open architecture. It uses the latest ITS technologies to provide real-time transportation system control, monitoring, and information. ATMS can detect the traffic situation and transport to traffic control center by the Internet (Rose et al., 1990). It can collect and manage Figure 22. Operation process between message set and data dictionary 287 Simple Transportation Management Framework Table 4. ParameterType Type Length (byte) MessageType ObjectName ObjectID Range globalDeviceSysTime globalDeviceSysTime ROCYear DeviceROCYear Integer 1 0~99 Month deviceMonth.0 Integer 1 1~12 Day deviceDay.0 Integer 1 1~31 Wekday deviceWeekday.0 Integer 1 1~7 Hur deviceHour.0 Integer 1 0~23 Config Search Search Response Activily Response 12H 0F+12 2BH 0F+42 14H 0F+C2 0F+02 0F+41 0F+C1 Minute deviceMinute.0 Integer 1 0~60 Second deviceSecond.0 Integer 1 0~60 DeviceState globalDeviceState HWReset deviceHWReset.0 Constant 1 1 10H 0F+10 CommReset deviceCommReset.0 Constant 1 1 11h 0F+11 LoopTestParameter deviceLoopTest.0 Integer 32 0~255 2CH 0F+47 2CH 0F+47 2EH 0F+C7 DbLockState deviceDbLockState.0 Integer 1 0~2 21H 0F+16 0F+46 0F+C6 EquipmentId deviceEquipmentId.0 Integer 2 0~ffh 18H 0F+40 29H 0F+C0 Password devicePassword.0 Char 6 ASCII 0F+45 0F+C5 DeviceEvent globalDeviceEvent RestartEvent deviceRestartEvent RestartMonth deviceRestartMonth.0 Integer RestartDay deviceRestatDay.0 Integer 1EH 0F+00 traffic situation to improve traffic efficiency and safety. The major services of ATMS consist of Network Surveillance, Probe Surveillance, Surface Street Control, Freeway Control, HOV Lane Management, Traffic Information Dissemination, Regional Traffic Control, Incident Management System, Traffic Forecast and Demand Management, Electronic Toll Collection, Emissions Monitoring and Management, Virtual TMC and Smart Probe Data, Parking Facility Management, 288 1FH 0F+15 1 1~12 1~31 Reversible Lane Management, and Road Weather Information System. ATMS Data Dictionary Figure 22 shows the operation process between message set (IEEE Std 1488- 1999, 2000) and data dictionary (IEEE Std 1489-1999, 1999). It is based on Open Systems Interconnection (OSI) communication protocol and supplies the common data format defined between applications. It can ensure the communication of systems. The OSI Simple Transportation Management Framework Reference Model is an abstract description for layered communications and computer network protocol design. It was developed as part of the OSI initiative. As shown below, in its most basic form, it divides network architecture into seven layers which, from top to bottom, are the Application, Presentation, Session, Transport, Network, Data-Link, and Physical Layers. ATMS C2F Level Definition C2F uses the communication protocol standard of SNMP and STMP. The messages must be defined as data objects and dynamic objects based on ASN.1. Example of Global Object Definition Using global object can setup time, database status on devices, for example, devices could be setup month, day, weekday, hour, minute, and second. reFerences Aidarous, S., & Plevyak, T. (1994). Telecommunications Network Management into the 21st Century. New York: IEEE Press. Perkins, D., & McGinnis, E. (1997). Understanding SNMP MIBS. Upper Saddle River, NJ: Prentice Hall, Inc. Rose, M. T. (1990). The Open Book: A Practical Perspective on OSI. Upper Saddle River, NJ: Prentice Hall, Inc. Stallings, W. (1993). SNMP, SNMPv2 and CMIP The Practical Guide to Network Management Standards. Reading, MA: Addison-Wesley Publishing Company, Inc. Stallings, W. (1996). SNMP, SNMPv2 and RMON. Reading, MA: Addision-Wesley Publishing Company, Inc. NTCIP Standard 1103 (2005). TS 3.2-1996 NTCIP Simple Transportation Management Framework - Amendment 1. ITS Standards Outreach, Education and Training Program, Institute of Transportation Engineers Management and Operations of Intelligent Transportation Systems (2003). ITS Standards Overview. IEEE Std 1489-1999 (1999). IEEE Standard for Data Dictionaries for Intelligent Transportation Systems. Feit, S. (1995). SNMP: A Guide To Network Management. New York: McGraw Hill, Inc. IEEE Std 1488-1999 (2000). IEEE Standard for Message Set Template for Intelligent Transportation Systems. ITU-T X.680-X.690 (1994). ISO/IEC 8824: Abstract Syntax Notation One (ASN.1). Steedman, D. (1993). ASN.1: The Tutorial & Reference. London: Technology Appraisals Ltd. National Highway Institute (2001). Using the National ITS Architecture for Deployment. 289 290 Chapter 17 Vehicular System Management Architecture and Application Platform Teng-Wen Chang National Taiwan University of Science and Technology, Taiwan, R.O.C. Jiann-Liang Chen National Taiwan University of Science and Technology, Taiwan, R.O.C. absTracT Notably, not all telematics services can be used in telematics terminals as a result of the varied platform standards. The main issues are that most telematics technologies depend on vertical, proprietary and closed per-OEM Original Equipment Manufacture (OEM) platforms, forming islands of non-interoperable technology and preventing third-party service providers from creating valuable services. In this study, the Open Gateway Service Initiative Vehicle Expert Group (OSGi/VEG) was integrated into an Android platform to generate a vehicular Android/OSGi platform that has the advantages of both original platforms, such as remote management, rich class sharing, proprietary vehicular applications, security policies, easy management of application programming interface (APIs), and an environment with increased openness. Furthermore, this study integrates the cloud computing mechanism into the Android/OSGi platform, which allows service providers to upload their telematics bundles onto storage clouds via the provisioning server. inTroducTion Over the past few years, with the enormous market potentials of telematics industry and the rapid development information technology, automotive telematics has been a booming area and indispensable technology, additionally, it also become a hot R&D area in mobile computing and ITS (Intelligent DOI: 10.4018/978-1-60566-840-6.ch017 Transport Systems). Telematics is the convergence of telecommunications and information processing for automation in cars. So far, quite a number of telematics services have been developed by automakers and third-party service providers, such as monitoring, emergency roadside assistance, navigation, diver aids, remote diagnostics, entertainment, web browsing, and so on. But not all of telematics services can be deployed into telematics terminal as a result of various standards of platforms. Be- Copyright © 2010, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global is prohibited. Vehicular System Management Architecture and Application Platform sides, the telematics market is immature now. The main critical issues are most of telematics technologies depend on the vertical, proprietary and closed per-OEM platforms, forming islands of no-interoperable technology and preventing third-party service providers from creating value added services. Consequently, numerous vehicle groups have been some working in establishing and developing open/standard embedded platforms for vehicles, these platforms contain OSGi/ VEG, AUTOSAR, AMI-C, CVIS, OSEK/VDX, Android, and so on. Figure 1 illustrates the open Linux operating system is ported into embedded on-board terminal, which not only provides a variety of device drivers such as CAN\LIN\FlexRay car buses, outnetworks connection modules and so on, and also offers resources management. The open/standard telematics platforms in telematics middleware layer mainly standardize telematics API’s and graphic/vocal HMI (human-machine interface), so that both service providers and car manufacturers can quickly deliver solutions on time to potential market and to simplify complexity of development. Besides, if service providers want to remotely deploy telematics services to on-board terminal, or road-side centers need to diagnose the situations of vehicular devices or set-up configurations of telematics applications in terminal, they should use remote management services rely on open/ standard telematics platforms. Telematics applications can be divided into four categories, including VANET embedded system, vehicular multimedia embedded system, intelligent driver aids embedded system, and Urban Nomadic/Pedestrians Telematics embedded system. The first part of VANET system which makes vehicle can communicate with other vehicles or road-side units via DSRC/IEEE 1609, for example if at front of vehicle have accident, it would broadcast emergent message to back vehicles based on VANET embedded system. The second part of multimedia system which makes drivers can be able to watch DVB programs, play on-line games, surveillance home conditions, and so on. The multimedia embedded system should take advantages of high-performance graphic SoC or 3D engines when it performed complicated operations. The three part of driver aids system which ensures security of drivers, economizes the use of power and so on. The final part of Urban Nomadic/Pedestrians Telematics system which provides interactive services between users and service providers in non-automotive environment, for example how to find neighbor locations of filling stations in urban. The rest of available telematics scenario is that vehicular drivers can download/upload or share information to telematics information platforms in Web servers based on out-networks connection capability. In this chapter, we will introduce how to combine OSGi/VEG into Android platform, making new vehicular Android/OSGi platform has both advantages of original platforms, such as remote management/deployment, rich class-sharing, proprietary vehicular applications, security policies, and so on. relaTed works google android open platform Overview of Android Platform The Android™ delivers a complete set of software for mobile devices: an operating system, middleware and key mobile applications. The Windows Mobile and Apple’s iPhone now provide a richer, simplified development environment for mobile applications. However, unlike Android, they’re built on proprietary operating systems that often prioritize native applications over those created by third parties and restrict communication among applications and native data. Android offers new possibilities for mobile applications by offering an open development environment built on an open source Linux kernel. Real hardware can be 291 Vehicular System Management Architecture and Application Platform Figure 1. Telematics architecuture accessed through series of standard API libraries, such as to manage the Wi-Fi, Bluetooth, and GPS devices. As Figure 2 illustrates the Open Mobile Alliance (OHA) and Google had supported Android platform, and hope to reach the goal of ensuring global mobile services to interoperability across devices, geographies, service providers, operators, and networks. At this stage, Google had released the open source of Android platform, providing the opportunity to create new adaptive mobile platform interfaces and applications designed to look, feel and function exactly as our imagination. Consequently, the Android platform had been ported in mobile devices, such as notebook, PDA, and automotive system in recently. In automotive system fields, the Intel and Wind River Systems companies are actively working on getting Android-powered infotainment operating system integrated into vehicles. in runtime, and native or third-party applications in application layer. The Android software stack components are as follows: • • Android Software Stack The Figure 3 illustrates the Android software stack, which composed of Linux kernel, a collection of Android libraries, application framework that provides management of android application 292 • Linux kernel: Core services (including device hardware drivers, process and memory management, security, network, and power management) are managed by a Linux 2.6 kernel. The kernel also provides an abstraction layer between the hardware and the rest of the stack. Libraries Running on top of the kernel, Android includes various C/C++ core libraries such as libc and SSL, as well as: ◦ A media library for playing multimedia resources. ◦ A Surface manager to provide display management ◦ Graphics libraries that include SGL and OpenGL for 2D and 3D graphics ◦ SQLite for data storage ◦ SSL and WebKit for integrated web browser and Internet security Android runtime: Including the core libraries and the Dalvik virtual machine, the Android runtime is the runtime environment for android application to be run in normal. Vehicular System Management Architecture and Application Platform Figure 2. Open environment for android platform • Core libraries: While Android development is done in Java/C++, Dalvik is not a standardized Java VM. The core Android libraries provide most of the functionality • available in the core Java libraries as well as the Android-specific libraries. Dalvik virtual machine: Dalvik is a register-based virtual machine that’s been Figure 3. Architecture of android software stack 293 Vehicular System Management Architecture and Application Platform • • optimized to ensure that devices run multiple instances efficiently and stable. It relies on the Linux kernel for threading and lowlevel memory/process management. Application framework: The application framework provides the classes used to create Android applications. It also provides a generic abstraction for hardware access and manages the user interface and application resources. Application layer: All applications, both native and third party, are built on the application layer using the same API libraries. The application layer runs within the Android runtime using the classes and services made available from the application framework. Features of Android Platform The unique features which other mobile platforms don’t own as follows: • • • Figure 4. Location-based services for android platform • • 294 Google map of navigation applications:Figure 4 illustrates Android platform provides the MapView widget lets navigation display, manipulate, and annotate a Google Map within your Activities to build map-based applications using the familiar Google Maps interface. Background service and applications: Background services let android applications that use an event-driven model, working silently while other applications are being used and in foreground. All applications are created equal: Android does not differentiate between the core applications and third-party applications. They can all be built to have equal access to a platform’s capabilities providing users with a broad spectrum of applications and services. P2P interdevice application messaging: Android offers peer-to-peer messaging mechanism that supports presence, instant messaging, and inter-device/inter-application communication. Breaking down application boundaries: Android breaks down the barriers to building new and innovative applications. The user can combine information from web to his/her individual’s Android platform, such as the user’s contacts, calendar, or geographic location to provide a more relevant user experience. Besides, all of the thirdparty applications can be downloaded to user’s Android platform from Android Market on free. Vehicular System Management Architecture and Application Platform open services gateway initiative (osgi) service platform Overview of OSGi Framework The OSGiTM specifications define a standardized, components-oriented, computing environment for networked services that is the foundation of an enhanced service-oriented architecture (SOA). Besides, the OSGi specifications were initially targeted at residential internet gateway with home automation applications (Lee et al., 2003; Myoung et al., 2005); however, the OSGi features of the standard and extensible that making enterprise to regard OSGi technology as key solutions. For example, Nokia and Motorola drove the OSGi technology standard for the next generation of smart phone, and the vehicle industry adopted the OSGi specifications by embedding them into the global system for telematics (GST) specifications, which is supported by many car manufactures. The OSGi framework provides general-purpose, secure support for deploying extensible and down- loadable java-based service applications known as bundles. OSGi bundles are applications packaged in a standard Java Archive (JAR) file, containing a manifest file that describes the relationships among bundles and OSGi framework, a series of compiled Java classes, and native code. Figure 5 illustrate the OSGi framework run on top of a Java virtual machine, and the framework provides a shared execution environment that installs, updates, and uninstall bundles dynamically without restarting the system. OSGi-Based Telematics Gateway The OSGi Alliance whitepaper had pointed out that the research and development of OSGi have extended to any networked environment, such as automobile, which provided some complex embedded devices and the requirement of deployable services for those devices. The OSGi framework make it possible to access, download and use services in automotive mobile environment, creating an enormous potential for automo- Figure 5. Architecture of OSGi framework 295 Vehicular System Management Architecture and Application Platform tive manufacturers and service providers to offer new proactive services to drivers and passengers (Gun et al., 2001). The open standard-based, service-oriented automotive infrastructure for the telematics architecture(Li et al., 2005; Zhang et al., 2004), including OSGi gateway, service infrastructure and self-automobile. The central control point of the service infrastructure, which is the OSGi-based telematics gateway, can be interconnected with various internal and external networks and buses. It acts as the execution environment for mobile automotive services. The following bundles provide great deal of critical services for OSGi-based telematics gateway: • • • 296 Monitoring bundle: It is the one of basic intelligent vehicle functions. Though monitoring state and performance of local vehicle devices, driver can acquire information about vehicles, such as temperature, and pressure of types, engine levels, etc. Navigation bundle: It can provide navigation system that offers navigational assistance to drivers. The system receives GPS position information signals which are processed to determine current position latitude and longitude coordinates, moreover the direction of travel. Communication bundle: It can provide OSGi-based telematics gateway to connect with in-vehicle or out-vehicle network. Besides, this bundle can make OSGi-based gateway to be extended its functionality by remote deployment of OSGi bundles that service provider provided. Through CAN, LIN or MOST(Zhou et al., 2006), the OSGi-based telematics gateway realizes local area connection among in-vehicle devices, and by using GPRS, GSM or even special ITS FM channels, it realizes wide area connection among vehicles, home gateway, remote service center, and roadside systems. • • • Driver aid bundle: It can provide driver assistance for vehicles, and offering some warning message to driver by lane recognition. Diagnostics bundle: It can provide remote diagnostic services, and complicated diagnostic information also can be analyzed and stored in remote systems, in this technique is similar with MYCAREVENT project (Weiss et al., 2006), making the telematics gateway to have the capacities of automatic repair and self-management. Information/entertainment bundle: At present, video and audio equipments have been widely applied in vehicles. Consequently, existence of researches that will shift to provide an infotainment server system for the in-vehicle users such that the network-enabled information appliances (IA) can access the information and perform the entertainment services from infotainment server (Hsu et al., 2005). The automotive operating system is also more significant to the vehicles (Ai et al., 2007), by supporting specific device drivers such as CAN/ LIN buses, which is used to communicate to other electronic control units (ECU) nodes for diagnostic purpose, consequently existence of researches that embedded OSGi platform to the vASOS automotive operating system (Sun et al., 2007). Figure 6 illustrates the OSGi-based middleware run on top of the K virtual machine (KVM), which is a compact, portable Java virtual machine intended for small, resource-constrained devices. In addition, those researches choose OSGi R3 implementation of Oscar, to develop its application bundles, such as navigation and CAN/LIN/MOST access bundle. In our studies, we had took advantage of OSGi R4 implementation of Apache Felix to integrate with Google Android platform, making a variety of Google APIs to turn into OSGi bundles, such as location-based, peer-to-peer communications, and Vehicular System Management Architecture and Application Platform Figure 6. Architecture of vASOS and OSGi-based middleware for vehicles network manager, etc. Not only supplying service providers to deploy its telematics services, but also enhancing Android runtime layer performance and remote management functionality. The goal of the Vehicle Expert Group (VEG) is to tailor and extend the OSGi specification in order to meet vehicle-specific requirements. To achieve this, the VEG will define a list of topics that cover vehicle-specific issues. Together with the automotive industries, the VEG will specify corresponding telematics API’s, so that both service provider and car manufacturers can quickly deliver solutions on time to market and to increase customer loyalty. The OSGi-VEG organization has the following automotive projects: • 3GT: The 3GT project is to help establish OSGi-based in-vehicle telematics platforms on the European mass market by ensuring interoperability between the products of different middleware providers, terminal manufacturers and service providers. This will be done by establishing common telematics interfaces for OSGi-based • • service delivery. More specifically, 3GT will develop OSGi-based specifications for the interface between vehicles and control centers as well as for the interface between control centers and service providers. ITEA EAST EEA: This automotive project is funded by the European Union and consists of major European automotive manufacturers, first-tier suppliers and research departments. The goal of EASTEEA is to enable hardware and software interoperability of in-vehicle ECU nodes through definition of an open, middlewarebased architecture. Stadtinfokoeln: Stadtinfokoeln is a Cologne-based project, funded by the German Ministry of Education and Science (BMBF), and focused on the delivery of parking services to the inhabitants or visitors of Cologne, Germany. The goal is to satisfy customer demand and to reduce traffic, which is related to the availability of parking space. An important role for this is the rapid development of the information and 297 Vehicular System Management Architecture and Application Platform • Communication-Technology in the mobile environment. To fulfill all requirements the decision was made to use an OSGi service platform framework on the vehicle platform. This enables the rapid development and deployment of new services. TOP-IQ: The Eureka-project TOP-IQ (Hackbarth, 2003) focuses on the development of a new generation of on-board computers (OBC) for luxury cars and trucks. The OBC facilitates telematics services for transport, logistics, offers the management, billing, and delivery of new services such as online navigation, trailer management/ tracking and tracing of cargo, etc. Safe Deployment of OSGi Bundles The Security of deployment vehicular services is crucial issues in telematics environment, supposing vehicle owners would like to install third-party applications to their on-board vehicle system, but if those vehicular applications have many bugs, such as sending mass messages, accessing the security sensitive parts of vehicles, and causing Figure 7. Weaving process in vehicular environment 298 high-cost of consumption, they maybe encounter the hazardous conditions. Therefore, in open environment it is hard to establish meaningful trust relationships, and even when one can, trust is not equated with quality. The existence of research which utilizes the Aspect-Oriented Programming (AOP) to implement security enforcement policy to weave into OSGi bundles (Phung et al., 2008). Additionally, the AOP provide the cross-cutting functionalities into complicated vehicular applications, and program concerns in a software system can be captured and encapsulated in to so-called aspects. Figure 7 illustrates the weaving process scenario in vehicular environment, which consists of in-vehicle system, control center, and third-party service providers. The vehicle owners would like to install vehicular application bundle by making a corresponding request in the GST standard. Before installing this application bundle on the in-vehicle system through wired/wireless network, the control center weaves the application with defined aspect such as security policies, so that the woven bundle can be operated with enforcement security policies into the in-vehicle system. Vehicular System Management Architecture and Application Platform comparions oF osgi and android Figure 8 illustrates both OSGi and Android define a development platform for developing serviceoriented computing mechanism in mobile devices, and existence of similar VM technology, service model, component model and feature of open. But the main difference is that Android platform utilizes process-based isolation to enhance resource management, however OSGi platform takes capability of classloader-based isolation model, which allows OSGi bundles to selectively export some of their internal packages to other required bundles. Bundles can declare their package dependencies and the framework is responsible for managing dependency relationship between required bundles and provided bundles. Based on classloaderbased isolation mechanism, OSGi framework has advantage of rich class sharing. On the other point of view, Android platform uses background services to perform time-consuming operations, but usually these services are heavyweight, and only can be accessed by utilizing inter-process communication, which makes service access slower and more expensive. In OSGi, service components are lightweight. They are called directly when finding them in service registry. Consequently, OSGi services play significant role in keeping components lightweight and loosely coupled. In OSGi specification, it defines many of remote management services, such as SNMP, CMISE, CIM and OMA DM (Hyun et al., 2008). By utilizing remote management services, we can easily remote-manage OSGi framework in many situations, but these services are non-available in Android platform. Android platform lack of class-sharing ability seriously limits the sharing of components and remote device management. The lack of versioning and lack of management APIs is a further limitation. Presumably, OEMs will provide some solutions for these important requirements. open embedded android/osgi plaTForm For TelemaTics location of osgi Framework in android platform The Dalvik VM use the device’s underlying Linux kernel to handle low-level functionality including security, threading, process, and memory management. It’s also possible to read/write C/C++ applications that run directly on the underlying Figure 8. Comparisons of OSGi and android 299 Vehicular System Management Architecture and Application Platform Linux OS. But Dalvik VM merely can execute Dalvik executable files (e.g., classes.dex), a format optimized to ensure minimal memory footprint; consequently it can’t perform general Java packages as well as OSGi bundles, so that OSGi framework operations to fail when Android realizes OSGi functionalities in runtime. Besides, as Figure 9 shows, if the OSGi framework success to run in Android Runtime layer, it can’t directly interact with users, because users only can catch the sight of Android Activities (same as application) in application layer. For this reason, the Android/ OSGi Activity must to grab the report or log message of OSGi framework, and should to implement Android UI to present information to the users. We consider the OSGi R4 implementation of Apache Felix to integrate with Android platform, because this container is more portable and light-weight, and being appropriate for embedded hardware device such as telematics system. android/osgi conversion mechanism The Android/OSGi conversion mechanism can assume OSGi framework to dynamically loading bundles through Android Davik VM. In first, as the Figure 10 shows the Android/OSGi conversion mechanism, which can make OSGi framework perform a series of complied Java classes from OSGi bundles in Android runtime layer, by means of classes.dex file to referencing OSGi bundle’s general Java classes. Therefore, the classes.dex file can be regarded as bridging role between OSGi framework and bundles. We had implemented Android/OSGi conversion code through Dalvik.system.dexFile API, and importing two parameters, such as package name(bundle API packages) and ClassLoader (reading the Java classes) objects, and then conversion code will return the collection of Java classes, so that OSGi framework can obtain this collection objects to indirectly perform the functionalities that OSGi bundles provided. 300 Figure 9. Location of OSGi framework Figure 10. Android/OSGi conversion mechanism Vehicular System Management Architecture and Application Platform After accomplishing the Android conversion mechanism, we will make sure all OSGi bundles or general Java packages to contain classes.dex files, so that OSGi framework can assume any bundle to work, making API packages to share, and letting consumer bundles to query the OSGi services that producer bundles provided. We had adopted the DEX conversion tool that Android SDK provided. This tool can easily make Java complied file to have classes.dex file that Dalvik VM is able to execute. Figure 11 illustrates the conversion flow of OSGi bundles to become apk file (Android executive extension); in originall, we had made the OSGi bundle projects to be compiled, and if no exception will generate a series of Java compiled classes. After making complied classes to come into being jar file, we took DEX tool to make jar file containing classes.dex, and then by using aapt command to transform OSGi bundles into Android apk files. In the end, by using adb push command to push converted apk files to Android platform. With Finishing Android conversion mechanism and converted OSGi bundles, we had ported OSGi framework to Android underlying layer, and other functional bundles such as telnet, deployment admin, http, and remote manager bundles etc, so that Android platform can be assisted and managed in remote site that original platform doesn’t have this ability. Figure 12 shows the script of starting OSGi framework in Android runtime layer, by means of Dalvik VM reading the main class of OSGi framework, afterward will enter OSGi console mode, and showing list of some installed bundles. Nevertheless, in this technique, the users can’t directly interact with OSGi framework in real hardware device, so we had implemented Android/OSGi Activity and GUI interface, to acquire corresponding OSGi information from underlying layer, and also embedding OSGi framework into Android application that can enhance it to be more adaptive and powerful. development of android/ osgi activity We had created Android/OSGi Activity in Android application layer, which can fetch the OSGi corresponding information from Android runtime layer, and providing interactive environment between users and OSGi framework. In first integration of procedure, we must to implement Android Acitivity interface, letting our Android/OSGi application to have Android Activity properties. Unlike most traditional environment, Android applications have no ability of control over their own life cycles, Instead, Android application framework must to be in charge of android applications state, and react accordingly, taking particular care to be prepared for untimely termination. When Android/OSGi Activity becomes active state and need to operate continually, the Android applica- Figure 11. Conversion flow of OSGi bundles to become apk files 301 Vehicular System Management Architecture and Application Platform Figure 12. OSGi framework run in android runtime layer tion framework will pay attention to ensure that the Android/OSGi Activity remains responsive. Therefore Android application framework has the capacity of monitoring platform resources at any time, if necessary, will kill some empty or background processes to free resources for highpriority applications. The Android application is called Activity, which represents a screen that an application can present to its users. Android Activity typically includes at least a primary interface screen that handles the main UI functionality for our application. This is often supporting additional by secondary Activities for entering information, and providing different perspectives on our data. For example, we had created the Android views of OSGi console, making users can enter OSGi commands and catch the sight of OSGi corresponding Information. Every Android Activity has an initializer method called onCreate(), by using this initializer method, we can distribute the memory to perform the launcher thread which can make OSGi framework to be started. Figure 13 illustrates the cross-communication operation that Android/OSGi Activity launch the FelixService (Android background service) which can represent the GUI interface of OSGi console and to load 302 the main class of OSGi framework after Android/ OSGi Activity onCreate() method had already been initialized. Android services run without a dedicated GUI, they usually silently perform background works that users can’t perceive. Before Android services to run, they must to be attached with our Android/OSGi Activity. From different point of view, the Android/OSGi Activity provides Android Context object, which makes Android application framework to be able to manage this Activity life states, and communication between Android/OSGi Activity with other native or thirdparty Activities. To ensure that our Android/OSGi Activity remains responsive, we move all slow, and timeconsuming operation off the main Activity thread and onto a child thread, that is continually reading the main class of OSGi framework, and updating GUI interface when OSGi console view has been changed or GUI bundles provide the new screen services. Consequently, we had implemented the Felix service to do the above-mentioned operations, as Figure 14 shows that Felix service can directly find out the location of OSGi framework in Android runtime layer, and extract the main jar file of OSGi framework, which is Apache Felix of R4 OSGi implementation, and by means of Vehicular System Management Architecture and Application Platform Figure 13. Cross-communication operation Dalvik ClassLoader object to read the main class continually in child thread, so that we can start OSGi framework in Android application layer. In addition, we also had implemented the OSGi console view that can fetch the OSGi corresponding message, and represent to users. The OSGi console view has two input/output interfaces that make users to immediately interact with OSGi framework. Figure 15 illustrates two-way Communication Mechanism between OSGi and Android. The Android/OSGi Activity(described in Figure 13) can transmit these received information to the OSGi framework after finishing above-mentioned operation(First step of Figure 15, sending data such AP table), so that OSGi framework can regard these information as permanent properties data (key/value pairs) stored in OSGi environment(Step 2 in Figure 15), and OSGi bundles can retrieve these properties through preference service to implement advanced applications such as network manager, and navigation bundle, etc(Step 3 in Figure 15). In the same way, when OSGi bundles becoming ACITVE state, it also can transmit some explicitly or implicitly intents to launch useful Activities. Consequently, on the strength of this two-way communication mechanism, we can make Android applications to connect with OSGi framework and bundles easily. embedding osgi Framework to android application In traditional, the Android/OSGi platform utilizes the OSGi services and service registry as the extensibility mechanism. This mechanism will let any application to run completely on top of OSGi framework, but this is not always possible, in other words, the components of system should follow the OSGi standard to be implemented, which will result in increasing the development times and costs. Besides, it should spend extra-time to communicate with OSGi framework in runtime layer. Therefore, if any Android application wants to use either the OSGi services or provided APIs by bundles, it needs complicated Android/OSGi communication mechanism. We had created extender mechanism that makes OSGi framework to tightly embed into Android applications in application layer as Figure 16 shows, in this way, any android application can host the instance of OSGi framework by utilizing reflection approach, and application can externally load the services which OSGi bundles provided. Nevertheless, the extender mechanism has some drawbacks about lack of dynamic changes in OSGi bundles/services and configuration of OSGi instance. Therefore, we had combined two potential mechanisms with 303 Vehicular System Management Architecture and Application Platform Figure 14. Inter-architecture of the Felixservice Figure 15. Two-way communication mechanism between OSGi and Android Android platform via cross-communication interface between felixService and OSGi Framework, management between framework layer and embedded OSGi Activity, to provide more integrated Android/OSGi environment. The Felix framework instance doesn’t utilize ServiceReference array object to get referenced services, but on the contrary it use Service Tracker 304 interface to monitor services. As a result of utilizing traditional service-oriented mechanism, the producer bundles should use its BundleContext object to register services in service registry, and if services referenced by consumer bundles had been changed, we must implemented ServiceChanged event to be careful of no exceptions had been generated, hence this mechanism is more Vehicular System Management Architecture and Application Platform Figure 16. Tightly integrated Android/OSGi environment inconvenient than service tracker mechanism. The service tracker takes advantages of whiteboard pattern, making embedded OSGi framework to utilize server bundle to monitor referenced services changes in central way. For example, as Figure 17 shows, if referenced services had been modified, the server bundle automatically callback modifiedService method to rebind these services without implementing ServiceChanged event handler. For example, the embedded OSGi Activity has capacity of tracking UI services, and if UI services have existed, the embedded OSGi Activity would utilize child thread to call setContentView method to update the foreground screen to users. Vehicular android/ osgi applicaTions Vehicular Testbed It is inconvenient and high-cost to test our Android/ OSGi applications to real vehicles, consequently as Figure 18 illustrates we build a vehicular testbed to test and verify the functionalities of Android/ OSGi applications. Besides, we regard sensorbased LEGO robots as simulated vehicles, and vehicular Android/OSGi platform acts as on-board terminal for simulated vehicles. Instead of utilizing car buses, this on-board terminal mainly controls simulated vehicles through Bluetooth like as IBM Figure 17.Service tracker of embedded OSGi activity 305 Vehicular System Management Architecture and Application Platform iDrive systems. And communications can be done via Wi-Fi among different vehicles. android/osgi applications We developed various telematics services to establish intelligent vehicles in mobile environment, which mainly consist of line follow, object detection, keep distance, and so on. As Figure 19 shows the first part of application is making vehicles to follow guideline to ensure security of drivers and no necessary to finding your path Figure 18. Vehicular testbed Figure 19. Android/OSGi applications 306 in order to lighten the use of power. The second part of application is making vehicles to have visual intelligence that can identify objects like as traffic signals. The rest part of application is making vehicles to keep distance with front/ back vehicles. Android/OSGi applications can be remote-deployed into on-board terminal by automaker or service providers, but if some of applications access the significant components of vehicle, the vehicular Android/OSGi platform can enforce security policies to avoid accidents through AOP-based OSGi weaving mechanism. Vehicular System Management Architecture and Application Platform Original Android platform can make developer to create map-based Activities using Google Map as a user interface element. We have full permissions to access the map, include change the zoom level, move the centered location, use overlays, provide map-contextualized information and functionality, and so on. But if service providers want to design navigation applications based on map-based Activities, they must to create many heavyweight services to perform time-consuming operations. Besides, Android platform lack of class-sharing among different applications and islands of no-interoperable processes, which often limit usage of system resources and remain non-responsive. Figure 20 illustrates vehicular Android/OSGi platform has feature of lightweight applications, which application size is less 4KB than original application. But vehicular Android/ OSGi platform must to spend about 3 seconds to start OSGi framework in first time, and additional memory size of OSGi framework. remote management Figure 21 illustrates vehicular Android/OSGi platform mainly provide three management consoles, which include telnet, web management, and remote deployment console. The first part of telnet console can make remote manager to diagnose/ manage Android/OSGi applications though telnet interface, which extends telnet bundle that Oscar R3 implementation provided. Original Android platform can’t utilize telnet services to diagnose/ manage its conditions of applications or system configuration. The second part of web console management can make remote manager to diagnose/manage Android/OSGi applications based on web techniques. The web management console has been established rely on underlying OSGi bundles, which contain http service, declaractive service, log service, configadmin, deployadmin, metadata, and so on. Although original Android platform has http service, that can give users Figure 20. Lightweight Android/OSGi applications 307 Vehicular System Management Architecture and Application Platform Figure 21. Remote management of vehicular Android/OSGi platform way to surf internet depend on WebKit browser, it doesn’t have web server, which allow remote users to login/logout into platform. The rest part of remote deployment console can make applications be able to be deployed into remote vehicular Android/OSGi platform. This way is similar with Android market deployed mechanism, but has something differences, such as original Android platform can’t administrate the identity of users and automatically validate security and stability of applications. 308 reFerences Ai, Y., Sun, Y., Huang, W., & Qiao, X. (2007). OSGi based integrated service platform for automotive telematics. In Proceedings of IEEE International Conference on Vehicular Electronic and Safety, (pp. 1-6). Gun, N., Held, A., & Kaiser, J. (2001). Proactive services in a distributed traffic telematics application. International Workshop on Mobile Communication over Wireless LAN: Research and Applications. Vehicular System Management Architecture and Application Platform Hackbarth, K. (2003). OSGi – Service-DeliveryPlatform for Car Telematics and Infotainment Systems. Advanced Microsystems for Automotive Applications 2003 (pp. 497-507). Berlin: Springer. Phung, P. H., & Sands, D. (2008). Security Policy Enforcement in the OSGi Framework Using Aspect-Oriented Programming. In Proceedings of the IEEE International Conference on Computer Software and Applications, (pp.1076-1082). Hsu, R. C., & Chen, L. R. (2005). Integrated Embedded System Architecture for In-Vehicle Telematics and Infotainment System. Proceedings of the IEEE International Symposium on Industrial Electronics, 4, 1409–1414. Sun, Y., Huang, W. L., Tang, S. M., Qiao, X., & Wang, F. Y. (2007). Design of an OSEK/VDX and OSGi-based embedded software platform for vehicular applications. In Proceedings of IEEE International Conference on Vehicular Electronic and Safety, (pp.1-6). Lee, C., Nordstedt, D., & Helal, S. (2003). Enabling Smart Spaces with OSGi. IEEE Pervasive Computing, 2(3, July-Sept), 89-94. Li, Y., Wang, F., Feng, H., & Li, Z. (2005). OSGibased service gateway architecture for intelligent automobiles. In Proceedings of the IEEE International Conference on Vehicles Symposium, (pp. 861-865). Myoung, K., Heo, J., Kwon, W. H., & Kim, D. S. (2005, July). Design and implementation of home network control protocol on OSGi for home automation. In . Proceedings of the IEEE International Conference on Advanced Communication Technology, 2, 1163–1168. Weiss, E., Gehlen, G., Lukas, S., & Rokitansky, C. (2006). MYCAREVENT- Vehicular Communication Gateway for Car Maintenance and Remote Diagnosis. In Proceedings of the IEEE International Conference on Computers and Comunication, (pp. 318-323). Zhang, D., Xiao, H. W., & Hackbarth, K. (2004). OSGi based service infrastructure for context aware automotive telematics. In . Proceedings of IEEE International Conference on Vehicular Technology, 5, 2957–2961. Zhou, Y., Wang, X., & Zhou, M. (2006). The Research and Realization for Passenger Car CAN Bus. In Proceedings of the IEEE International Forum on Strategic Technology, (pp. 244-247). 309 310 Chapter 18 Remote Vehicular System Management Functions and Information Structure Teng-Wen Chang National Taiwan University of Science and Technology, Taiwan, R.O.C. Jiann-Liang Chen National Taiwan University of Science and Technology, Taiwan, R.O.C. absTracT Due to the rapid development of information technology, the network has already spread to every corner of vehicle. With all kinds of ECU devices appear in the vehicle, and it brings the more and more convenient living. On purpose solving heterogamous technologies that are incompatible with each other, developed a “WBEM-based Remote Management and Heterogeneous Vehicular Network Diagnosis System” on OSGi Gateway. This system can focus on a variety of problems come from vehicle network, and find out what are the problems or where are the problems happened. If the problem still can not be solved properly, we must to seek for help from remote managers. The users can acquire enough information without understanding how to control every device, so that the users can help near diagnosis system to solve vehicle network’s problems and to promote the abilities of near network diagnosis. inTroducTion Due to the vehicular network management system has not been standardized and lacking interoperability. In addition to the existing network management system has a simple structure consisting of server, provider and client, as communication between admin users and management server. This study DOI: 10.4018/978-1-60566-840-6.ch018 constructs a remote vehicular network management and diagnosis system with OSGi platform (Ai et al., 2007; Sun et al., 2007) and WBEM (Web-Based Enterprise Management) to speed up the realization of the heterogeneous home network. Usage of OSGi platform provides several benefits. The most important one that it integrate different types of vehicular networking technologies, because currently most available vehicular network middleware and technology have no compatibility with each Copyright © 2010, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global is prohibited. Remote Vehicular System Management Functions and Information Structure Figure 1. Remote managing protocols other, home devices based on these heterogeneous middleware cannot communicate with one another, even though they are physically connected. The OSGi service platform is initially chosen for its capabilities to integrate components and services from different providers in the heterogeneous home network. (Marples et al., 2001; Saito et al., 2000) The OSGi service platform is specifically designed for devices that can manage remote devices through remote managers. Figure 1 illustrates these devices that need to be established some of remote managing protocols, such as SNMP, CMISE, CIM, OMA DM, and more. The OSGi Alliance decided that no managing protocol can be preferred over others because no protocol is suitable for all cases. The OSGi Alliance now has been working to develop a standardized management API to build functionality of remote management into unattended devices. This is a very powerful concept that offers the same interoperability as a standard protocol. However, the benefits of this concept are not always immediately obvious. The benefit of a standard protocol is that any device can be managed by any manage operators. Consequently, we apply Web-based Enterprise Management (WBEM) managing protocol to vehicular network environ- ment. Using WBEM protocols, remote managers can easily manage devices through web interfaces, and operate with any manufacturer’s device, regardless of the underlying protocols. The proposed system uses the WBEM to provide more effective resource management and a larger range of services than before. Because the different types of vehicular network technologies and management instruction, need the different technology to manage the each vehicular network, like SNMP, Telecommunications Management Network (TMN) and so on. In order to overcome the situation that WBEM and CIM (common information model) provide an excellent management environment and promotes the information exchange across a variety of underlying technologies and platforms supported interoperability. The WBEM offer extreme flexibility and efficiency to manage method, that communication with each other different management type. Figure 2 illustrates overview of integrating OSGi platform with WBEM these two technologies and Remote management through Web-based Enterprise Management Architecture to the vehicular gateway. The WBEM-based Remote Vehicular Network Management and Diagnosis System can figure out the status fast and easily, and realize the benefits 311 Remote Vehicular System Management Functions and Information Structure Figure 2. Integrating OSGi platform with WBEM of integrating these two technologies, significant research has been carried out the telematics in the near future. relaTed works We make a brief introduction of vehicular network system which are popular nowadays, and make use of scenario situation to state vehicular network environment in the future, state a lot of technology applied to vehicular network in this chapter. Besides, related work would also be discussed. Vehicular network environment Figure 3 illustrates the vehicular network environment, which can be divided into two categories, including in-vehicle network architecture and out-vehicle network architecture. In the in-vehicle network, the vehicle might include four kinds of car bus protocols such as CAN (Controller Area Network), LIN (Local Interconnect Network), MOST (Media Oriented Systems Transport) and FlexRay (Mariño et al., 2009). The CAN bus protocol is main backbone for in-network environment in current, and it has the goal of making vehicle more reliable, safe and fuelefficient while decreasing wiring harness weight and complexity. Besides, CAN bus protocol has 312 applied widespread in industrial automation and automotive/truck applications. The FlexRay protocol provides high-speed, deterministic and fault tolerant communications for in-network. It also can be compatible with existing networks, such as CAN, LIN, and so on. But above-mentioned protocols can’t intercommunicate to each other. Consequently, the OSEK/VDX operating system acts a role which makes the message communicate between two different car protocols (Kim et al., 2007). In the out-vehicle network, the OBU (On Board Unit) in the vehicle can communicate to infrastructure via out-network modules, or connect with another vehicles via DSRC/IEEE 1609. The remote home service provider and remote vehicular service provider can provide its particular services to automotive user. open services gateway initiative (osgi) service platform The OSGi™ Alliance was founded in March 1999. Its mission is to create open specifications for the network delivery of managed services to local networks and devices. The OSGi organization is the leading standard for next-generation Internet services to homes, cars, mobile phones, desktops, small offices, and other environments. With the advent of ubiquitous communication, PCs as well Remote Vehicular System Management Functions and Information Structure Figure 3. Vehicular network environment as diverse devices, such as sensors and information home appliances, are now being linked together, often using heterogeneous communication protocols. In addition, broadband access has become much more widespread and it is common to see homes and offices with always-on connections to the Internet. Against this backdrop, increased attention is being paid to gateways that provide key capabilities for functional interworking between devices and portal capability for using services offered by external networks, including the Internet. The OSGi Service Platform is delivered in many Fortune Global 100 company products and services and in diverse markets including enterprise, mobile, home, telematics and consumer. The core component of the OSGi specifications is a framework. It provides a general-purpose, secure, and managed Java framework that supports the deployment of extensible and downloadable applications known as bundles. OSGi-compliant devices can download and install OSGi bundles, and remove them when they are no longer required. The Framework manages the installation and update of bundles in an OSGi environment in a dynamic and scalable fashion. To achieve this, it manages the dependencies between bundles and services in detail. It provides the bundle developer with the resources necessary to take advantage of Java’s platform independence and dynamic code-loading capability in order to easily develop services for small-memory devices that can be deployed on a large scale. The OSGi framework architecture has a number of layers as depicted as Figure 4. web-based enterprise management (wbem) WBEM stands for Web-Based Enterprise Management. WBEM is an ongoing initiative started by industry leaders such as Compaq, Microsoft, Cisco, Intel and over 70 others. This initiative proposes a common method of managing enterprise systems. It merges existing management solutions with the latest advances in Web technology. WBEM is a set of Internet standards which gives the ability to interconnect between different management standards and environments. WBEM allows managing both software (operating systems, applications) and hardware (computers, network 313 Remote Vehicular System Management Functions and Information Structure Figure 4. OSGi framework architecture devices) by creating a common layer which unifies and simplifies management through WBEM compliant applications. WBEM is an initiative of DMTF and it includes a set of technologies that enables the interoperable management of an enterprise network. The DMTF has developed a core set of standards that make up WBEM, which includes a data model, the CIM standard; an encoding specification, CIM-XML encoding specification; and a transport mechanism, CIM operations over HTTP. Figure 5 illustrates the relationships among WBEM Standard Technologies architecture. The CIM specification is the language and methodology for describing management data. CIM is an object-oriented schema for modeling the objects. The CIM schema can be divided to three areas; the core model, the common model and the extension model. First, the core model captures notions that are applicable to all areas of management. Second, the common model is an information model that captures notions that are common to a particular technology. For example, it includes the model for systems, applications, networks and devices. Last, the extension model represents technology-specific extensions of common models. Figure 5. Relationships among WBEM standard technologies 314 Remote Vehicular System Management Functions and Information Structure The CIM-XML encoding specification defines XML elements, written in Document Type Definition (DTD) that can be used to represent CIM classes and instances. The CIM operations over HTTP specification defines a mapping of CIM operations onto HTTP that allows implementations of CIM to interoperate in an open, standardized manner and completes the technologies that support WBEM. Therefore, in the WBEM architecture, the management information is described by the CIM schema, converted to XML, and finally embedded in an HTTP payload to transport to the target node. Figure 6 illustrates the WBEM architecture (Thompson et al., 1998), which includes a WBEM Client, and WBEM Server. WBEM Server has CIM Object Manager (CIMOM) which is a central component that routes information about objects and events between components. The CIMOM responds to the operations defined in the “CIM operations” specification such as create, modify, and delete. It also checks the syntax and semantic of the messages, and provides security. Providers are so-called instrumentation agents. Namely, Providers actually obtain the information from the resources and forward it to the CIMOM. A WBEM Client is commonly represented as the management application, and it can get the information by sending a request message to the CIMOM instead of directly accessing the providers. Figure 7 shows the various layers through which an operator’s request passes and through which the response returns when a managed object is accessed. A command from the operator or higher-level management system originates with the application logic which handles graphical screens, interpretation of command-line interfaces, etc. The command is then mapped to the abstract object model. Fundamentally, the Object Abstraction layer has to know that, when the operator enters a command to create a new OSPF service, for example, then this means the creation of instances of various CIM classes. This knowledge may Figure 6. WBEM Architecture be hard-coded, table-driven or driven from the model. Once the actual CIM commands have been determined, they are encapsulated in CIM-XML and passed to the HTTP client. This is responsible for any negotiation required with the WBEM server and the correct transfer of the request. Addition to traditional CIM operations using XML, CIM operations using Binary XML (Park et al., 2006) enhance a performance of message transportation between WBEM clients and a WEBM server. Once the request reaches the WBEM server, it is passed from the HTTP server to have the CIM-XML reconstituted into CIM commands. The CIM Object Manager (CIMOM) examines these and determines how they should be handled: by the CIMOM itself, with reference to the schema stored in the repository, or by a provider. If it is passed to the provider, then the provider accesses the real device or service to retrieve the requested information or carry out the requested configura- 315 Remote Vehicular System Management Functions and Information Structure Figure 7. WBEM Client/WBEM server interactions tion. The response from the provider follows the reverse path back to the operator. WBEM is architecture, not an implementation. Implementations exist though. Sun Microsystems has released a Java WBEM and implemented CIMOM with Java Management Extensions and Microsoft has its CIMOM implementation called Windows Management Instrumentation (WMI). Another similar SDK that facilitates development of WBEM servers is the SNIA CIMOM, a free product with open source code that is compatible with the Sun WBEM SDK API. Table 1 shows existing WBEM implement. The WBEM technology has the following features: • • 316 Real Standard: Back by many industry, like Microsoft, IBM, Sun, HP, and so on. Mature implementations exist: Reduce development and maintenance effort. • • • Standard Interface for different monitoring systems: Much fabric management uses the WBEM standards. Standard Interface for information providers: Future APIs available for simplicity. Can replace log file style providers: Provide CIM repository. Replacing existing management standards such as SNMP or CIMP is not WBEM attempt to, but to provide a framework embracing existing management standards and protocols. This would allow the integration of distributed management services provided by different management platforms and applications. Consolidate and unify the data provided by existing management technologies is the key purpose of the WBEM initiative. Remote Vehicular System Management Functions and Information Structure Table 1. Existing WBEM implement Producer or initiator Platform Programming language License Elements and capabilities Java 1.2 or higher SISSL (Sun Industry Standards Source License) CIMOM, Client API, Provider API, MOF Compiler, CIM Workshop, providers for Solaris OS (only in Solaris) WBEM Services Sun Microsystems Solaris 8 (Spark/ Intel) open source: platform-independent (Java) Pegasus The Open Group (IBM, Compaq, HP, BMC) Linux, Unix (AIX, HPUX, Solaris), Windows NT/2000/9x C++ Open Source CIM Object Broker (CIMOM), Client API, Provider API SNIA WBEM Storage Networking Industry Association (SNIA) platform-independent (Java) Java (since 1.1.8) SNIA Public License Open Source OpenWBEM Caldera Unix, Linux, Solaris, other C++ Open Source CIMOM, Client API, Server API, MOF Compiler, WQL (WBEM Query Language) tools SBLIM IBM Linux C++ (Native Provider Interface) Common Public License a set of providers for Linux system Integrated with MS Windows operating system API (WBEM Provider DLL), functionality based on the operating system kernel General Public License libCIM, PaulA (CIMOM API) WMI Microsoft MS Windows 98/2000/XP any language capable to use DLL API B4WBEM B4WBEM Linux Perl simple network management protocol (snmp) The Simple Network Management Protocol (SNMP) is an application layer protocol that facilitates the exchange of management information between network devices. Absolutely this protocol is used in the IP-based home network environment. SNMP enables network administrators to manage IP-based home network performance, find and solve network problems. SNMP is a request-reply protocol running over UDP through TCP operation is possible. SNMP is extensible, allowing vendors to easily add network management functions to their existing products. SNMP is not a mere paper specification, but an implementation that is widely available today. SNMP is based on a manager and agent model. An SNMP-managed network consists of three main components including: managed devices, agents and Network-Management Systems (NMSs). Figure 8 illustrates the relationships of these three components model. These managed objects are arranged in a virtual information database known as a Management Information Base (MIB). The ‘manager’ is the console where network management functions are performed, and the ‘agents’ are the entities or software modules that interface with the actual devices being managed. A Management Information Base (MIB) is a collection of information which defines the properties of managed objects. Managed objects are comprised of one or more object instances, which are essentially variables. The SNMP manager and 317 Remote Vehicular System Management Functions and Information Structure Figure 8. NMS, agents and managed devices model for SNMP agent use a MIB and relatively small set of commands to exchange information of the managed devices that contain network nodes or managed objects. An object identifier or object ID uniquely identifies a managed object in the MIB hierarchy. The MIB hierarchy can be depicted as a tree with a nameless root, the levels of which are assigned by different organizations. For example of the Figure 9 illustrates the MIB tree. The top-level object IDs represent different organizations, while the lower-level object IDs are allocated by associated organizations. The diagram below illustrates an MIB tree where the top-level object IDs are different organizations, and the lower-level object IDs are the associated object. To identify the ‘system’ illustrated in the diagram, a unique ID for the object can be ‘iso.org.dod.internet.mgmt. mib-2.system’ or the equivalent numeric object descriptor, ‘1.3.6.1.2.1.1’. Addition to IP-base network, as PLC networks and their applications grow with the advances in PLC technologies, major PLC chipset and modem vendors are trying to provide network management capabilities in their devices by defining their own private management information base (MIB). Some PLC management researches (Park et al. 2008; Park et al. 2008) focus on integrated management of multi-vendor PLC networks by using SNMP technology. 318 wbem compared to snmp In this section, WBEM will be compared against SNMP. Table 2 shows the different between WBEM and SNMP. As for the conventional network management protocols, the Simple Network Management Protocol (SNMP) is the most popular among the Internet Protocol Based (IP-BASED) local area network, while no obvious Figure 9. Example of the MIB tree with various hierarchies Remote Vehicular System Management Functions and Information Structure winner among other types of networks. However, SNMP is not a suitable protocol for remote home network diagnosis system because (1) the polling mechanism used by SNMP may cause network congestion in a large network or crossing the boundary of wide area network (WAN), and SNMP places the burden of data collection entirely on the management side, (2) SNMP agent cannot provide the historic record of an equipment or data, and (3) SNMP cannot use a unified data descriptor to preserve the flag, state and configuration of all the managed equipments. Also has follow reasons besides these. The conventional SNMP network management approaches have several difficulties and challenges (Chen et al., 1998). First is incompatibility and platform-dependence in managing heterogeneous networks. Second is high investment. Comparatively, Web-Based management approaches provide the following advantages: lower cost, scalability, platform-independence, more flexibility and remote access, and easier dynamic information. managing archiTecTure and diagnosTic meThod For remoTe conFiguraTion oF heTerogeneous local neTworks As the communication and network technologies progress rapidly, the number of network-related facilities used in the home and small business is also increasing. The raising of the information appliances enables the popularity of easy-to-use, convenient, and Internet connectable appliances in the home and small business networks. In addition, the co-existence of different heterogeneous inter-connection technologies, such as IEEE 802.3, IEEE 802.11, Bluetooth, IEEE 1394, power line, home plug, and so on, is becoming a common phenomenon. In complicated and heterogeneous network environment, the management of the home and small business network is both a harsh challenge and an urgent demand for the users, appliance designers and service providers. In general, when a home network or a small business network encounters a Table 2. Comparison between WBEM and SNMP Features SNMP WBEM Data encoding format Binary encoding (BER) XML (xml/CIM) Data description model MIB (Management Information Base) MOF (Managed Object Format) Namespace Defined by OID (ObjectIDentifiers) Defined by CIM Schema Data structure Vendor-based CIM-based object model Transportation protocol UDP HTTP/TCP Protocol encapsulation IP,UDP, SNMP IP, TCP, (SSL, SHTTP), HTTP, XML Reliability UDP is connectionless by TCP functions Protocol messages types SNMPv1: Get, GetNext, Set, Trap SNMPv2: all from SNMPv1, GetBulk, Inform, Reply SNMPv3: all from SNMPv2, new Trap and Report SimpleReq, SimplerRsp, MultiReq, MultiRsp Security SNMPv1, SNMPv2 - only community name SNMPv3 - USM (User-based Security Model HTTP 1.1 SSL+HTTP SHTTP Management flexibility Low, generally network devices only High, management model is device independent Other protocol interconnectivity Low may be realized by proxy agents. Mapping to many protocols: SNMP, DMI, CMIP itd (Mappings are described by DMTF standards) 319 Remote Vehicular System Management Functions and Information Structure problem, the user usually does not have sufficient knowledge or expertise to perform diagnosis or trouble-shooting. Therefore, this service provides a good business opportunity of potentially lucrative revenue for the telecommunication or Internet service provider (ISP) companies. As a result of automotive environment is more complicated and changeable than home environment, we take our managing system into home network, to validate and perform the managing functionalities of proposed architecture. The proposed managing architecture and diagnostic method for remote configuration of heterogeneous local networks, which includes at least one sub-network agent, a local area network (LAN) management module and a remote LAN module, as shown in Figure 10. Each sub-network agent manages its sub-networks via its own management protocol, and collects the sub-networks’ information. The LAN management module receives the requests from heterogeneous local networks via these sub-network agents, and converts the information associated with each request into a common information module (CIM) to seek a solution for each request. The remote LAN module receives the unsolved requests from the LAN management module via a channel, configures the heterogeneous local networks and uses compatible interface at a remote side to accomplish the management and diagnosis for the heterogeneous local networks. Figure 11 show the module details of the proposed architecture. In following section, we detail the proposed architecture of remote managing of heterogeneous local area network. sub-network agents Each sub-network agent manages the sub-network through its own management protocol, and collects sub-network information. The LAN management module is coupled to the sub-network agent and a cross-internet channel respectively, receives one or more requests from the heterogeneous local network through the sub-network agent, and converts the information accompanying the Figure 10. Architecture of remote managing of heterogeneous local area network 320 Remote Vehicular System Management Functions and Information Structure Figure 11. Module details of the proposed architecture node information inside the sub-network, information between nodes, and other related network information. Each sub-network agent uses its own managing protocol, such as SNMP or common management information protocol (CMIP), to obtain the sub-network information. Each sub-network build-in, customize or filter out the analysis of the event relevance. Each sub-network agent also communicates with LAN management module through protocol conversion module, and receives the instruction from LAN management module to control related sub-networks. The information after sub-network agent filters or integrates is returned to LAN management module so that remote LAN module knows the information and the status of each sub-network. local area network (lan) module request into a common information model to find a solution for the request. The remote LAN module receives the requests, which the LAN management module fails to solve, through the channel, accesses and configures the heterogeneous local network remotely, and uses interface compatible with the heterogeneous local network to realize the management and diagnosis of the heterogeneous local network. Sub-network act as drivers and interface between the LAN management module and the Sub-network. Sub-network agent collect information of sub-network, and warn about or respond to the related event automatically. The information provided by the sub-network agent includes the The LAN module provides internet connection to the remote LAN module through a channel, and receives one or more management and network information of one or more heterogeneous local networks through the sub-network agent and converts into a common information model for providing the remote LAN module a unified management information and operation. When the heterogeneous local network encounters problems, the LAN management module searches for solution first, and when the LAN management module fails to provide solution, the LAN management module requests the remote LAN module or a remote manager for assistance so as to realize the management and diagnosis of the heterogeneous local network.LAN management module includes a LAN control module, a network information module, a network protocol conversion module, a common information storage media, a user interface, and a security module. Network protocol conversion module is responsible for communicating and coordinating with sub-network agent through related protocol and mechanism to obtain the basic and management information of sub-network. Network 321 Remote Vehicular System Management Functions and Information Structure protocol conversion module issue a service, such as through registration, to notify other related services of the service content so that other related services use the new service, such as through installation pr-executing instruction. Network protocol conversion module converts the obtained information into a common information model or similar format. The common information model is an overall management information model describing all the computer systems and network equipments in an enterprise network environment, including a set of specifications and a set of schemas. Network protocol conversion module transmits the converted information to network information module. The data conversion is such as converting the sub-network instruction into common information model. The functions of network protocol conversion module could be added flexibly. Network information module analyzes and organizes the converted common information from network protocol conversion module, and then selects and defines a common information model suitable for storing network information for storing in common information storage media. Network information module is responsible for accessing information stored in common information storage media required by LAN control module for diagnosis and management, and receives the issued diagnosis and managed instructions and passes the instructions to network protocol conversion module for performing diagnosis and management. The common information model has sufficient expressive capability to represent all the managed objects, and has sufficient expansion capability to accommodate new managed objects as well as accessing the management information effectively. For example, the common information model object manager (CIMOM) of the management infrastructure of web-based enterprise management (WBEM) plays the role of network information module. Common information storage media is the actual information storage media to match network 322 information module. Common information storage media is a database or a specific format file, such as management object format (MOF) file matching CIM in the management infrastructure of WBEM. The information stored in common information storage media is only accessed through network information module. LAN control module is the control center of LAN management module, providing basic network management functions, network diagnosis process, algorithm for solving network problems, network information update process, and communicating with remote LAN control module. The basic network management functions include numerous items such as allowing the user to know through user interface the network internal basic information, such as network topology, network traffic, network speed, and node information, or even the software installation on each node. In the problem diagnosis, the collected information is used to analyze the possible cause of the problem, and forwards the problem to LAN control module. If LAN control module fails to solve the problem internally, a request is sent to remote LAN module for assistance to provide a solution. When the problem is solved, LAN control module or remote LAN control module report to the user or warn the user for preventing similar events in the future. When the problem is not completely solved, LAN control module or remote LAN control module will also inform the user of the problem handling status. For example, through the node analysis to obtain the node equipment information, LAN control module or remote LAN control module use the information to search for equipment manufacturer for repairing and post the information on the user interface to inform the user. The cause or the solution to the problem is recorded on common information storage media or remote common information storage media for future reference. User interface show the network information, such as network topology, each node information, network traffic, and network speed. Through the Remote Vehicular System Management Functions and Information Structure graphic and web-based interface showing the network status, the user easily use the mouse or button to process complicated network problem and issue management instructions with simple and clear guiding mode. For the user, the interface is simple and effective. Therefore, the user understands the network internal basic information through the use of the user interface. Security module is responsible for related security mechanism, such as security authentication mechanism, data encryption mechanism, protecting the internal data access mechanism for LAN control module, and billing mechanism. Security authentication mechanism is to verify whether the user is the legitimate user or administrator so as to protect LAN control module from invasion by illegitimate remote LAN control module. During information transmission, a data encryption mechanism is provided to prevent data from theft. The data encryption mechanism is controlled by security module and remote security module, respectively. These two security modules both encrypt and decrypt the data transmitted and received on either side. The encrypted data will not be easily theft or utilized so as to achieve the security objective. remote lan management module Remote LAN management module receives the unsolved request from LAN management module through channel, such as request from sub-network, remotely access and configures the heterogeneous local network, and adopts an interface compatible to the heterogeneous local network to realize the management and the diagnosis of the heterogeneous local network. Remote LAN module includes a remote LAN control module, a remote network information module, a remote common information storage media, a remote user interface, and a remote security module. Remote LAN control module, remote network information module and remote common information storage media of remote LAN module basi- cally operate in a way similar to those modules in LAN management module. The difference lies in remote LAN control module is designed for remote LAN diagnosis and management service center. Therefore, remote control module monitor, manage, and register LAN, such as all home network and small business network. In general, remote LAN module collects and utilizes global data, such as periodically obtaining the topology of each LAN. In performing LAN management and diagnosis, remote LAN control module will request LAN control module of a certain LAN management module for detailed information, and stores the information in remote information storage media for assisting the solving of the problem that LAN control module of LAN management module fails to solve. algorithms A diagnosis method for remote configuration of heterogeneous local networks comprises the steps of: detecting through at least a sub-network agent whether any sun-network connecting to the sub-network agent encounters any problem; when detecting at least a sub-network encountering problem, collecting the management and network information of the sub-network with the problem, and transmitting to a LAN management module; converting the management and network information of the sub-network with the problem into a common information model through the LAN management module, and performing diagnosis to determine whether a solution can be provided; when the LAN management module fails to provide a solution, requesting a remote LAN module or a remote management for assistance to realize the management and diagnosis of the sub-network with the problem; and responding the diagnosis message to a user interface or the sub-network agent. Figure 12 illustrates diagnosis methods for remote configuration of heterogeneous local networks. 323 Remote Vehicular System Management Functions and Information Structure A method for updating network information for a management architecture for remote configuration of heterogeneous local network, comprising the steps of: determining whether network information is periodic information or dynamic information; if periodic, using the extract instruction to obtain the update information; if dynamic, collecting and processing the network information and converting the processed network information into a common information and obtaining the update information; filtering the updated periodic or dynamic information and determining whether to store remotely; if to store remotely, informing a remote LAN module and storing the updated information to a remote common information storage; and if not to store remotely, storing the updated information to a local common information storage. Figure 13 illustrates network information update process. A method for adding a sub-network to a managing architecture for remote configuration of heterogeneous local networks, comprising the steps of detecting a new network through a network protocol conversion module; determining the attributes of the new network through a LAN management module; when the LAN management Figure 12. Diagnosis methods for remote configuration of heterogeneous local networks 324 Remote Vehicular System Management Functions and Information Structure Figure 13. Network information update process module fails to provide the support service to the new network, sending a request to a remote LAN module; the remote LAN module finding a suitable sub-network agent to the new network and transmitting to the LAN management module; and when the new network having a sub-network agent, performing the registration and announcement of the new network. Figure 14 illustrates process of adding a new sub-network. Vehicular sysTem archiTecTure Figure 15 show the system architecture. The system architecture contains two parts. The left part is OSGi platform and the right part is WBEM server. The OSGi platform contains some bundle build upon OSGi framework; include authentication bundle, vehicular network diagnosis bundle, IP network agent bundle, Power Line network agent bundle, IEEE 1394 network agent bundle and connect interface bundle. The bundles further describe as follow: 325 Remote Vehicular System Management Functions and Information Structure Figure 14. Process of adding a new sub-network • • forwards the problem to vehicular network diagnosis bundle. If vehicular network diagnosis bundle fails to solve the problem internally, a request is sent to remote manager for assistance to provide a solution. Connect interface bundle: Connect interface bundle is adapter between OSGi platform and WBEM server. All information collected by network agent will convert to CIM model by connect interface bundle and store on WBEM repository or pass to remote center. Connect interface bundle also receive the instance from WBEM server. Authentication bundle: Authentication bundle is to verify whether the user is the legitimate user or administrator so as to protect home server from invasion by illegitimate remote client. The remote client provides a user interface that manager could watch the vehicular network information, manage device or sent diagnosis information to local home server if local don’t have enough solution. wbem-based remote Vehicular network management and diagnosis system • • 326 Network agent bundle: The IP network agent bundle collects information of IP network and uses SNMP managing protocol to manage device. Other network agent do same thing but using its own managing protocol. Vehicular network diagnosis bundle: Vehicular network diagnosis bundle providing basic network management functions, network diagnosis process, algorithm for solving network problems. In the problem diagnosis, the collected information by network agent bundle is used to analyze the possible cause of the problem, and Figure 16 shows the WBEM-based Remote Vehicular Network Management and Diagnosis System GUI. The WBEM-based Remote Vehicular Network Management and Diagnosis System GUI contains three parts: the left panel shows the vehicular network topology by using tree view; the top-right panel contains some buttons, and the bottom-right panel, which display messages about the detail device information and allow sent solutions info to WBEM-based Local Vehicular Network Management and Diagnosis System. The vehicular network topology is local vehicular network topology include IEEE 1394 network topology, IP-base network topology, and Lon- Remote Vehicular System Management Functions and Information Structure Figure 15. System architecture Figure 16. WBEM-based remote vehicular network management and diagnosis system GUI Works network topology. When click the device on left-panel topology, top-right panel will show the detail information of this device. For example in Figure 16, by clicking the Port7 device we could get the device description is Hardware x86, device uptime is 1 day 21:29:20.5, location is restroom, IP address is 192.168.1.50 and some other detail information show on top-right panel. All process in WBEM-based Remote Vehicular Network Management and Diagnosis System are matching management object format (MOF) file CIM in the management infrastructure of WBEM. The MOF of system includes attribute and method of system process. The lonworksdeviceinfo, ieee1394deviceinfo, and ipdeviceinfo are the defined network information model. Information 327 Remote Vehicular System Management Functions and Information Structure Figure 17. XML for invoking/return the snmpX method of each sub-network is described in this model. The topologyX is the method operation that client could get local sub-network topology. The snmpX is the method operation that client could get oid- value by input ipaddress and oid. The findip is the method that do mac to ip convert. In addition, the new CIM model could add to repository through OSGi update or download. Figure 18. WBEM-based Local Vehicular Network Management and Diagnosis System GUI 328 Remote Vehicular System Management Functions and Information Structure Figure 19. Diagnosing of the changes of SNMP switch network topology All information show on remote client is through WBEM server and is CIM model. It uses RMI or CIM Operations over HTTP to transmit information between the residential gateway and remote manager. The operations and messages are described abstractly, and then bound to a concrete network protocol and message format to define an endpoint. The CIM Operations over HTTP using an XML format for describing network services. Figure 17 illustrates invoked and returned operations for transmitting managing message. wbem-based local Vehicular network management and diagnosis system The WBEM-based Local Vehicular Network Management and Diagnosis System main GUI and OSCAR bundle list is shown in Figure 18. The GUI contains three parts: the top-left panel shows the vehicular network topology by using tree view; the top-right panel contains some buttons, and the bottom panel, which display messages about the home network situation or receive information from remote client. The client GUI shows IEEE1394 network topology, SNMP switch network topology, LonWorks network topology. In the IEEE1394 network topology that adopts the driver of the libraw1394 to get the network information because Java language doesn’t support the IEEE1394 drive. In the SNMP switch network topology adopts the SNMP4J API that classes are capable of creating, sending, and receiving SNMPv1/v2c/v3 messages. SNMP4J receives SNMP messages through the listen port of transport mappings to get the Ethernet network information. In the LonWorks network topology adopts the web page to simulate the i.Lon 100 web page. The i.Lon 100 web service can control the devices on and off or set up the time when the devices will be on and off, and it also shows the device status. When the topology changes, the text panel will show the possibility problems to let user fixed the problem easier. Figure 19 illustrates diagnosing process and the changes of SNMP switch network topology after plug in the IP-based device. That SNMP network topology is changed from three nodes into the four nodes. All process on WBEM-based Local Vehicular Network Management and Diagnosis System are matching management object format (MOF) file CIM in the management infrastructure of WBEM. No matter is topology information or diagnosis knowledge. reFerences Ai, Y., Sun, Y., Huang, W., & Qiao, X. (2007). OSGi based integrated service platform for automotive telematics. In Proceedings of the IEEE International Conference on Vehicular Electronic and Safety, (pp. 1-6). Chen, J. (1998). A Study of Web-Based SNMP Network Management with a Simple Java Applet Network Monitoring Tool. Department of Computer Science and Engineering Auburn University, Alabama. 329 Remote Vehicular System Management Functions and Information Structure Kim, J. H., Seo, S. H., Moon, T. Y., Hwang, S. H., & Jeon, J. W. (2007). A method of improving the reliability of the gateway system by using OSEK/ VDX. In Proceedings of IEEE International Conference on Control, Automation and System, (pp. 2328-2833). Mariño, P., Poza, F., Dominguez, M. A., & Otero, S. (2009). Electronics in automotive engineering: A top-down approach for implementing industrial field bus technologies in city buses and coaches. IEEE Transactions on Industrial Electronics, 589–600. doi:10.1109/TIE.2008.2002723 Marples, D., & Kriens, P. (2001). The Open Services Gateway Initiative: An Introduction Overview. IEEE Communications Magazine, 39(12), 110–114. doi:10.1109/35.968820 Park, C. K., Kang, J. M., Choi, M. J., Hong, J. W., Lim, Y. H., & Choi, M. S. (2008, Jan.). An Integrated Network Management System for MultiVendor Power Line Communication Networks. In Proceedings of the IEEE International Conference on Information Networking, (pp.23-25). Park, C. K., Kang, J. M., Choi, M. J., Hong, J. W., Lim, Y. H., Ju, S., & Choi, M. S. (2008). Definition of common PLC MIB and design of MIB Mapper for multi-vendor PLC network management. In Proceedings of the IEEE International Symposium on Power Line Communications and Its Applications, (pp.152-157). 330 Park, J. G., Ahn, C. W., Cho, H. N., Byun, I. S., Desmons, F., & Kim, S. W. (2006). A method for representing and transporting CIM operations using binary XML in the WBEM architecture. In Proceedings of the 8th International Conference on Advanced Communication Technology, (Vol. 3, pp.20-22). Saito, T., Tomoda, I., Takabatake, Y., Arni, J., & Teramoto, K. (2000). Home gateway architecture and its implementation. IEEE Transactions on Consumer Electronics, 46(4), 1161–1166. doi:10.1109/30.920474 Sun, Y., Huang, W. L., Tang, S. M., Qiao, X., & Wang, F. Y. (2007). Design of an OSEK/VDX and OSGi-based embedded software platform for vehicular applications. In Proceedings of the IEEE International Conference on Vehicular Electronic and Safety, (pp. 1-6). Thompson, J. P. (1998). Web-based enterprise management architecture. IEEE Communications Magazine, 36(3), 80–86. doi:10.1109/35.663331 331 Chapter 19 Using Wireless Mesh Network for Traffic Control Kun-Chan Lan National Cheng Kung University, Tainan, Taiwan, R.O.C. absTracT Wireless mesh networks (WMN) have attracted considerable interest in recent years as a convenient, flexible and low-cost alternative to wired communication infrastructures in many contexts. However, the great majority of research on metropolitan-scale WMN has been centered around maximization of available bandwidth, suitable for non-real-time applications such as Internet access for the general public. On the other hand, the suitability of WMN for missioncritical infrastructure applications remains by and large unknown, as protocols typically employed in WMN are, for the most part, not designed for realtime communications. In this chapter, we describe a real-world testbed, which sets a goal of designing a wireless mesh network architecture to solve the communication needs of the traffic control system in Sydney, Australia. This system, known as SCATS (Sydney Coordinated Adaptive Traffic System) and used in over 100 cities around the world, connects a hierarchy of several thousand devices -- from individual traffic light controllers to regional computers and the central Traffic Management Centre (TMC) - and places stringent requirements on the reliability and latency of the data exchanges. We discuss some issues in the deployment of this testbed consisting of 7 mesh nodes placed at intersections with traffic lights, and show some results from the testbed measurements. inTroducTion Adaptive traffic control systems are employed in cities worldwide to improve the efficiency of traffic flows, reduce average travel times and benefit DOI: 10.4018/978-1-60566-840-6.ch019 the environment via a reduction in fuel consumption. One of the main and most common functions of such systems lies in adaptive control of traffic lights. This ranges from simple lengthening or shortening of green and red light durations in an intersection according to the actual presence of cars in the respective lanes, to coordination of green light Copyright © 2010, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global is prohibited. Using Wireless Mesh Network for Traffic Control phases among neighboring intersections on main thoroughfares. This adaptivity is made possible with the use of sensors (typically in the form of magnetic loop detectors embedded under the road pavement) that feed data to roadside traffic light controllers, and a communications infrastructure that connects among the intersections and a traffic management centre, as well as, in some cases (typically in large cities), a hierarchy of regional computers (RC) that perform the control decisions for respective portions of the system. Traditionally, the communications layer of traffic control systems has been based on wired connections, either private or leased from public telecommunications operators. While for many years such leased lines (operating at 300bps) have served their purpose well, they have several shortcomings, such as a significant operating cost, inflexibility, and difficulty of installation in new sites. In certain cases, alternative solutions, operating over public infrastructure, have been deployed for specific sites where private or leased lines were not a viable option; these ranged from ADSL, regular dialup, or cellular (GPRS). However, using public network for traffic control could suffer from inconsistent delay jitters and reliability issues. For example, previous experimental studies (Chakravorty et al., 2002) have shown GPRS links could have very high RTTs (>1000ms), fluctuating bandwidths and occasional link outages. In recent years, there has been considerable interest in wireless mesh networks and their deployment in metropolitan areas, from both a commercial and a research perspective (Lundgren et al., 2006). Trials in several major cities in the US (e.g. Philadelphia, New Orleans, and others (Tropos networks, 2009; Locust world, 2009) and worldwide (e.g. Taiwan (Mobile Taiwan) have shown mesh networks to be a viable technology that can compete well with alternative “last-mile” connectivity solutions to the public. Correspondingly, most of the research on metropolitan-area wireless mesh networks (MAWMN) has focused on maximising the throughput that can be extracted 332 from them, in the anticipation that their major use will be public, for purposes such as accessing the Internet or conducting voice calls (Ganguly et al., 2006). On the other hand, little attention has been directed to the aspects of reliability and latency, which are most important if MAWMN are to be considered for replacement of mission-critical infrastructure, such as traffic control system communications. In this chapter, we describe a wireless mesh network testbed that has been built. The testbed physically covers seven traffic lights in the suburban area of Sydney. These intersections are chosen because they represent a typical suburban area with lots of traffic, foliages, pedestrians and high-rise residential buildings. In addition, the inter-node distance (ranging from 200 to-500m) is representative of 90% of the distance between traffic controllers in the Sydney CBD (Central Business District) area. The testbed nodes have been custom-built. The contribution of this paper is three-fold. First, to the best of our knowledge, our work is one of the first efforts to study the feasibility of using wireless mesh networking for traffic control. Second, we describe the details of our testbed implementation and some experiences we gained during the deployment of the testbed in an urban environment. Finally, we present some initial measurement studies of link characteristics of different wireless and wired technologies used in our testbed (including the use of 900MHz, 2.4GHz and 3.5GHz radios and Ethernet-over-powerline). Although our results are still very preliminary, they are useful to serve as a reality check toward the goal of applying wireless mesh networking to traffic control applications. The rest of this paper is structured as follows. In section 2, we describe the details of SCATS, the traffic control system used in Sydney and many other cities worldwide, and its communication requirements. We describe related work in Section 3. Section 4 presents a simple analysis of the topology of traffic lights in the Sydney met- Using Wireless Mesh Network for Traffic Control ropolitan area, and in particular the dependence of the degree of connectivity of the mesh on the radio range. Section 5 describes the details of our testbed implementation. We present some initial measurement results of link characteristics of different radio technologies used in our testbed in section 6. Section 7 discusses the experiences we gained during the deployment of our testbed. We conclude the paper and discuss the future work in section 8. scaTs oVer wireless In this section, we first describe the details of SCATS (Sydney Coordinated Adaptive Traffic System) and its communication requirements. We then discuss the benefits and research challenges when running SCATS on a wireless mesh network. The scaTs Traffic management system Developed and maintained by the Roads and Traffic Authority (RTA, formerly Department of Main Roads) of the state of New South Wales, the Sydney Coordinated Adaptive Traffic System (SCATS) is one of the most popular traffic management systems used worldwide. Its main task is to adjust, in real time, signal timings in response to variations in traffic demand and system capacity. Real-time data from traffic controllers are collected and transported to a central traffic management centre (TMC) for analysis and optimum control of road traffic. The performance of SCATS, therefore, depends critically on the capabilities of the underlying communication system that transports roadside data to and from the TMC. The existing communication system of SCATS relies strongly on third-party wired infrastructure (provided by Telstra, Australia’s largest telco). The bulk of the communications to the intersections, namely the traffic light controllers and vehicle detectors, are predominantly made using serial point-to-point connections over standard voicegrade telephone lines, using 300bps modems. This is also the most common method of connecting between the TMC and other lowbandwidth devices, including variable message signs, variable speed limits, ramp meters, and over-height detectors. At the core of the SCATS operation is a periodic exchange of messages between the controlling computer and each and every intersection (via the point-to-point links). This exchange happens every 1sec, and is initiated by the computer which sends to the intersection’s local controller a command message, instructing it about the next phase it should switch to and the timing of that switch. The controller, in turn, is required to reply with an acknowledgement, which includes information from the intersection’s sensors. If an acknowledgement is not received within 1sec from the time the command message is sent, it is retried once; after the second time an acknowledgement fails to arrive, the communications link is declared failed, and SCATS instructs all controllers at the respective cluster of intersections to fall back into a ‘default’ self-controlling mode, where decisions about the timing of green light phases are made locally and independently. Likewise, a controller will fall back to this mode upon not receiving a command message. Once triggered, a controller will stay in the self-controlling mode for at least 15 minutes; if another communications failure happens during this time, the duration of this mode will be extended by another 15 minutes, and so on. Obviously, the self-controlling mode, where the decisions at intersections are uncoordinated, can lead to a severely suboptimal traffic control, particularly in a busy thoroughfare during rush hour. Accordingly, though the bandwidth required from the communication links is quite low (comfortably handled by 300bps modems), the 1sec latency is critical for an efficient operation of the system. 333 Using Wireless Mesh Network for Traffic Control The currently used SCATS infrastructure, based on wired communications, suffers from the following problems: • • • Slow installation and inflexibility. In most cases, installing a new line at a road site (especially a remote site) involves earth excavation, which is very slow and with adverse effects on existing infrastructure. High capital and operating cost. The installation of a wired connection at a new site, or repairs at an existing one, carries a high cost due to the material and labour required. More importantly, the ongoing fees for leasing the wires from the telephone company run very high; currently, RTA pays nearly $40 million annually to Telstra in leasing fees for connecting the traffic signals and other roadside devices to SCATS. Low bandwidth. Modem-based leased lines support bandwidth less than 32 Kbps. While these low-bandwidth telephone lines are adequate for connecting traffic signal sensors, they cannot provide adequate support for connecting high-bandwidth applications, e.g. higher solution video cameras, which increasingly becoming necessary to effectively monitor traffic pattern on our roads. going wireless With wireless solutions, there is no cabling involved. Wireless can therefore provide fast installation and exceptional flexibility. Cost can be reduced significantly by building a private wireless network, because there will be no monthly charges to be paid to telephone company (some small license fee may apply). Moreover, the installation cost will be low because there will be no cabling-related labour. The cost issue is, in fact, the major concern for most road authorities. Finally, it should be noted that recent advances 334 in wireless technology provide bandwidth that is more than adequate for connecting many highresolution roadside cameras to SCATS. One possible option for going wireless is to build a dedicated RTA wireless network using widely available, standards based, low-cost wireless technologies, e.g. IEEE 802.11x and 802.16x. 802.11x equipment is cheaper, less complex, and operates entirely in the unlicensed spectrum (no licensing fee). On the other hand, 802.16x is more reliable (has multiple carrier frequencies to avoid interference), has longer range, and better features to cater for a diverse range of communication needs of future roadside equipment. In addition, it is possible to operate 802.16x in both license and unlicensed spectrums. Despite of its enormous benefits, there are several challenges when roadside ITS equipment is connected via wireless media: • • Latency. Wireless can potentially increase latency. For example, IEEE 802.11x, uses a common wireless channel (it is cheaper to share channel) among many contending devices causing potential conflict. To avoid such conflicts, some form of medium access control (MAC) is implemented by these technologies. MAC introduces some delay before data can be transmitted on the wireless channel. Reliability. Wireless signals are susceptible to interference from other signals in the vicinity operating in the same or adjacent spectrum. Given that ITS equipment is deployed in public area, such interference will be the norm rather than exception. Interference can corrupt messages transmitted over the wireless medium. Some frequencies do not work well (or at all) if there is no direct line-of-sight between the two communicating end points. In a dynamic context of public roads, roadside equipment may frequently face line-of-sight problems due to transient Using Wireless Mesh Network for Traffic Control • • obstructions, e.g. a high vehicle carrying a tall crane etc. Also in vehicle-to-roadside communications, a car in the near-lane may obstruct communication between a far-lane car and the roadside equipment. Temporary outages, i.e., periods when no wireless signal is available, therefore, is a real issue to deal with. Security. What makes wireless so vulnerable is the fact that the attacker does not have to gain physical access to the channel from any predefined access point. Roadside wireless components are well within the wireless range of passing motorists and pedestrians, which make them vulnerable to intrusion, denial of service, and other forms of security threats. Scalability. As mentioned earlier, wireless systems are sensitive to interference from other communicating devices operating in the vicinity. Additionally, if a common wireless channel is shared among all devices within a given area (cell), the MAC delay increases rapidly as the number of competing devices increases. Another scalability issue arises from the processing overhead that is required at a central radio base-station. The more remote radios there are in communication with the central radio, the more processing that must take place. The radio controller at the basestation will simply not be able to process all incoming radio signals if there are too many of them. relaTed work Roofnet (Bicket et al., 2005) is an experimental 802.11b/g mesh network built by MIT. Each node in Roofnet has an antenna installed on the roof of a building. Aguayo et al. (2004) analyzed the link layer behavior on the Roofnet testbed and described the impact of distance, SNR and transmission rate on the packet loss. While Roofnet’s propagation environment is characterized by its strong Line-of-Sight component, our work differs from the prior work in that our links are generally heavily obstructed 1. In addition, our planned deployment strategy is different from the unplanned topology in Roofnet. For example, our antenna is mounted at a height of about 4 meters from the ground. But the trees on the road are typically higher than 7 meters. Similar to our work, The WAND project (Weber et al., 2003) has built a multi-hop wireless testbed in the centre of Dublin. They have 11 nodes mounted on traffic lights along a 2km route in urban area. However, their topology is simpler than ours (i.e. a chain topology) and the measurements they performed on their testbed were relatively limited. TFA project (Camp et al., 2006) aimed to provide broadband access to low income community in Houston area via wireless mesh network technology. Their architecture consist of two wireless tiers: an access tier to connect homes, businesses, and mobile users to the infrastructure, and a backhaul tier to forward traffic to and from the wired entry point. Jardosh et al. (2005) discussed the correlation of link reliability with the frame retransmissions, frame sizes and data rate by collecting trace data from a structured 802.11b network during a international conference. They concluded that sending smaller frames and using higher data rates with a fewer number of frames improves the performance of congested network. All the previous studies have been centered around maximization of available bandwidth for non-real-time applications such as broadband access for the general public. On the other hand, this testbed is the first to focus on using wireless mesh networking for traffic control which places stringent requirements on the reliability and latency of the data exchanges. 335 Using Wireless Mesh Network for Traffic Control a simple analysis oF The sydney TraFFic lighT Topology This analysis was based on data provided by RTA, indicating the latitude and longitude of traffic controllers. Figure 1 shows points at each traffic controller location. As shown in Figure 1, there are around70 controllers in a 4km × 6km area. To understand the effect of radio range on the degree of connectivity when the traffic controllers are forming a mesh network, we calculate the shortest distance (assuming that the radio has a circular radio range and have perfect coverage in Figure 1.Location of traffic controllers Figure 2. Map of intersection locations 336 that range) between every pair of traffic controllers, and the output was then sorted so that for each controller, its neighbours were listed (from nearest to furthest) with the distance to each neighbour. We then processed this data to determine how many neighbours of each traffic controller were within a specified range (from 100m to 1250m). The results of this analysis are shown in Figure 3, which shows on the y-axis how many traffic controllers had 0, 1, 2, 3, . . . neighbours when a given radio range is assumed (x-axis). The results shown in Figure 3 provide a rough indication of what radio range is needed if we are aiming to interconnect a certain number of nodes with each node having a certain degree (number of neighbours within range). For example, if we seek to interconnect 90% of the nodes (accepting that some alternative technology may be needed for the minority 10% of nodes) and require that each node has three neighbours (to provide redundancy and hence fault tolerance), then we require a radio technology with range of at least 1km. Note that, while city environments may have large densities of traffic controllers in both (lat/long) dimensions, in suburban environments controllers (particularly those that are important to synchronise with communication links) often lie linearly along main arterial roads, with few controllers in close range Using Wireless Mesh Network for Traffic Control orthogonal to the main arteries. Neighbours that form a line would not provide the same level of fault tolerance as those that are better separated angularly around a controller, since the links are less likely to fail independently. TesTbed In this section, we provide the details of this testbed. We first describe the environment that the testbed is located. Next, the hardware used and the software installed on each of these nodes are discussed. environment The testbed is located in the Sydney CBD (Central Business District) area. Seven intersections are selected to deploy the testbed, as shown in Figure 2 (specifically, intersection number m1 to m7). A number of custom-build embedded PCs with multiple wireless interfaces are used. The nodes are mounted on the traffic lights at a height of about 2-3m from the ground, and distributed along the streets in the form of rectangle covering an area of 500 × 1000 square metres at a distance of 200-500m apart. None of the nodes is in a clear line of sights of its neighboring nodes. The node Figure 3.Numbers of neighbours within certain radio range of RTA controllers at intersection m1 is connected to a gateway node in University of Sydney. The streets where the network is deployed are about 10- 20m wide and surrounded by building at least two stories high. The majority of these buildings are made of concrete and steel that block the propagation of radio signals into the neighboring streets. All these streets have busy public traffic during business hours. Most of the vehicles on the street have a height of less than 2.5m. But some double-decker buses (such as Sydney Explorer Bus) or truck can have a height of more than 5m. channel characteristics Wireless channels can be coarsely characterized by its path loss exponent. Pathloss described the attenuation experienced by a wireless signal as a function of distance. The amount of variations in pathloss between similar propagation scenarios is called shadowing. In other words, shadowing represents the difference between the signal power at different points in the same environment with the same estimated pathloss. Prior study (Rapport et al., 1996) showed that shadowing can be modeled as a zero-mean Gaussian random variable. Specifically, one can predict the received signal power at a given distance d with the following formula:PdBm(d ) = PdBm(d 0) - 10a log 10(d d 0) + _ where a is the pathloss exponent, _ is the shadowing and d 0 is the reference distance. The prior work (Rapport et al., 1996) suggested that the pathloss can range from 2 to 5 for outdoor urban environment. Such physical level measurements are important for an efficient deployment (i.e. overestimating pathloss can result in overprovioning network while underestimating pathloss can produce disconnected network). By using linear regression, the environment of this testbed is found to have a pathloss a = 3.1 and shadowing _ = 7.2 . Note that the observed pathloss in this environment is significantly lower than the suggested 337 Using Wireless Mesh Network for Traffic Control urban pathloss of 4 in the literature (Rapport et al., 1996). hardware The hardware components used for the nodes in this testbed are off-the-shelf products including the following, as shown in Figure 4. All the components are mounted on two sides of a metal plate for easy maintenance (one can simply swap an old plate with a new plate when we want to upgrade the node). A custom-built enclosure is made to house this component plate. • • Motherboard: A VIA MB720F Mini-ITX motherboard featuring an 1GHz processor and 1GHz memory is employed as the core in our system. Storage: The traffic pole sometimes vibrates a lot due to the passing traffic. Since that our node is mounted on a traffic pole, instead of using a hard-drive, we employ a 2G USB flash drive for storing OS and data. Unlike a hard-drive, a flash drive does not have a high speed spinning platter and is less failure-prone. Figure 4. Hardware component 338 • Wireless interfaces: Each node has two wireless interfaces to connect to its neighboring nodes, as shown in Figure 5. To allow the testbed users to experiment with different radio technologies, two different radio frequencies are currently used on the testbed: 2.4GHz (802.11b/g) and 900MHz radios. Specifically, the nodes at intersection m2, m3 and m6 are installed with two 2.4GHz mini-PCI wireless cards from Ubiquiti (SR2). The nodes at intersections m1 and m5 are equipped with one 2.4GHz Ubiquiti SR2 card (with a transmission power of 400mW) and one 900MHz Ubiquiti SR9 card (with a transmission power of 700mW). Finally, the nodes at intersections m4 and m7 are installed with two Ubiquiti SR2 cards. One of these two SR2 cards is connected to a 2.4GHz-to900MHz converter (from RFlinx) to send 2.4GHz signal output by the wireless card on 900MHz band. Due to its better penetration rate for buildings and trees, theoretically the use of 900MHz radios could result in a better connectivity than 2.4GHz radios (i.e. 802.11x). Hence, 900MHz Using Wireless Mesh Network for Traffic Control • • radios are used at intersection pairs m1m5 and m4-m7. These two intersection pairs have a longer distance (i.e. 400m and 500m respectively) than the other intersection pairs. Back-haul connection: In addition to the two Ubiquiti wireless cards, each node is equipped an “Unwired” modem (Unwired wireless, 2009) to establish a back-haul link back to NICTA for the purpose of remote management, as shown in Figure 5. Unwired is a Sydney-based metropolitan wireless ISP. The Unwired modem uses a proprietary protocol but claims to be a variant of WiMAX and operates in a licensed 3.5GHz band. Ethernet router: A Linux-based Ethernet router (Diamond Digital R100) is installed in each node. We employ this Ethernet router for several purposes. First, it is used as an Ethernet switch to connect the motherboard and the Unwired modem (and any IP-based devices such as a camera in the future). The Unwired modem is connected to the WAN port of the router, thus the router get a public Internet IP address. The motherboard has an Ethernet connection with the router’s 4-port switch. Second, the Diamond Digital router has an USB port which allows the motherboard to have a serial connection with the router’s USB port through an USB-to-serial adapter. Being able to establish a serial link to the motherboard allows the user to remotely login into the serial console for troubleshooting when the Ubiquiti wireless interfaces are not responding. Third, given that the router is a Linux box itself (runs on openWRT), we can run all the existing software (e.g. we are currently running DNS, NTP and VPN clients on it). Finally, the Diamond Digital router has an 802.11 wireless interface which can be used as an alternative • • • • • • link to remotely connect the mesh node in addition to Unwired and Ubiquiti links. Power: As shown in Figure 4, an off-theshelf power-board (with surge protector and fuse) and a PC power-supply are used to provide the power to all the components in the node. The power-board takes a 240AC power from the traffic light. Antenna: Nodes on this testbed are all installed with omni-directional antennas due to the following Cost: An omni-directional antenna is typically cheaper than a directional antenna. In addition, for a node which has n neighbors, n directional antennas are needed. On the other hand, one omni-directional antenna per intersection is sufficient to cover all the neighbors. Mounting. The space on the traffic light for the mounting of antennas is quite limited. It is comparatively more difficult to mount a directional antenna on the traffic pole in practice. An 8dBi omni-directional antenna is used for the 2.4GHz wireless card and an 6dBi omni-directional antenna is used for the 900MHz wireless card. Weatherproof: The temperature in the summer can be above 40 Celsius degree in Sydney. The temperature inside the node can be even higher. As shown in Figure 4, to provide enough air circulation inside the node, many holes are drilled on the bottom of the enclosure and made some air louvres on the side. Two temperature-controlled fans are used in the node to dissipate the hot air out through the louvres. In addition, a roof is mounted on top of the enclosure to shield the enclosure from direct sunlight and rain. Remote recovery: Due to the fact that the testbed is deployed in an outdoor environment, it is time consuming to visit the nodes when something goes wrong. In addition, given that nodes are mounted on the 339 Using Wireless Mesh Network for Traffic Control Figure 5. Testbed topology traffic lights which are a public asset, visiting any node on the testbed required calling out the RTA maintenance crew to gain access to the node. Therefore, some means of remote recovery are necessary. Therefore, one wireless remote switch is installed on each node (runs in the unlicensed 433MHz band), which allows one to reboot the node on-site when accessing the node via the 2.4GHz or 3.5GHz links fails. The ultimate goal of this testbed is to control traffic lights using wireless mesh networks. A pair of power-over- Ethernet adapters (Netcomm NP285) is used to connect the node to a traffic controller board in the curbside housing through the powerline. The traffic controller board sends and receives data via a serial interface. Hence, a serial-to-IP conversion is required for the communication between the traffic controller and the testbed (which runs IP). The traffic controller board is mounted inside an embedded PC and connected to the embedded PC’s motherboard’s (VIA MB770F) serial port. A serial-to-IP converter software is written and run on the PC to 340 encapsulate the SCATS data from the serial port of the traffic controller board into an IP packet as well as to encapsulate the IP packet from the regional computer and send its payload to the serial interface. In order to connect the testbed to the regional computer, a gateway node at University of Sydney is deployed. The gateway node has a reasonable line-of-sight to intersection m1 and connects to the m1 node with a 802.11 link. Note that the Unwired links are not used to connect the regional computer (RC) to the testbed due to the consideration of reliability, latency and cost issues. More details about the characteristics of Unwired links are described in Section 6. The RC is connected to the gateway node via AARNet (Aarnet - australia’s research and education network, 2009). The round-trip delay between RC and the gateway is about 1.2ms, and the throughput is typically over 100Mbps. software The testbed node uses a custom-built Linux OS image that consists of the following components: Using Wireless Mesh Network for Traffic Control • • • • The image size is small enough to be fit into an USB flash drive. And run completely in RAM (1GB). This allows one to enable ’clean’ reboots uncontaminated by previous experiments. Some kernel modifications are added for various driver supports for USB, madwifi and PXE reboot. Grub is used and modified to activate the watchdog timer at the time of boot-loading so that the watchdog timer can be started before any program start. Watchdog timer is used to reboot the motherboard when the system fails. Various tools are used including timesync, Open- VPN, some network support tools and software from Orbit project (Raychaudhuri et al., 2005) in our image. The image is built to be Debian-based for the compatibility with Orbit software. The OS image is based on DSL-N (Damn small linux). DSL-N provides most of the software support we need out of the box. The default syslinux bootloader of DSL-N is replaced with grub. OML software (Singh et al., 2005) from Orbit project is used to build the measurement collection infrastructure for the testbed. Two security mechanisms are currently implemented. First, OpenVPN is used for the Unwired links to each of the mesh nodes. Second, ssh and friends are used on all network interfaces. In addition, root access is disabled on all the machines. link latency The round-trip delay increases as the number of hops increases on the 802.11 links. In addition, the variation also increases significantly when there are more hops. There is no strong correlation between distance and link latency though. As shown in Figure 6, the latency does not increase from 300m to 400m. However, the variation increase significantly as the distance increases. It is due to that there are more retries at 300m than at 400m due to different line-of-sight conditions. We next examine the efficiency of power line communication. As suggested in Figure 7, given a distance of 100m, the link latency of power line communication is excellent. The average round-delay is about 3.6ms and the variations are very small. In addition, the largest delay for such a distance is less than 8ms. As described in Section 5, the Unwired network is used to carry out our back-haul traffic. To understand the expected latency of running management traffic over the Unwired network, the round-trip delay to the mesh node is measured. As shown in Figure 8(a), the average delay of sending traffic over the Unwired network to the mesh node is about 400ms. However, there is a large variation (the delay can be as long as 3 seconds) and significant number of outages. The delay and outages over Figure 6. Effect of distance on round-trip delay link characTerisTics In this section, some results of measured link characteristics from the testbed are described. Specifically, some statistics of the wireless link performance are discussed in terms of round-trip delay, loss and throughput. Ping is used to measure the round-trip delay and iperf for the throughput measurement. 341 Using Wireless Mesh Network for Traffic Control Figure 7. Latency of powerline communication the Unwired network are mostly contributed by the wireless link between the mesh node and the Unwired ase station. As shown in Figure 8(b), the average delay of the Unwired wireless link is about 200ms. The large delay variations and significant number of outages suggest that a public-shared wireless network like Unwired is not suitable for operating SCATS traffic. losses As shown in Figure 9 and Figure 10, the packet loss seems to be distributed uniformly over time. Figure 8. Round-trip delay over the Unwired network 342 However, the loss becomes more bursty as the number of hops and distance increase. Note that Figure 9 and Figure 10 are based on the results from a low probing rate (i.e. one-packet-persecond ping). The loss pattern might change if we change the probing rate. In addition, we do not find there is a strong correlation between packet loss and the distance. The line-of-sight condition (which is location-dependent) plays a more important role on the packet loss. The use of 900MHz radio results in a much lower loss rate (0.5%) than 2.4GHz radio (20%), which is not surprising though since 900MHz radio has a better penetration rate than 2.4GHz radio. Finally, for a distance of 100m, the loss rate of power line communication is almost negligible (less than 0.1%). Throughput As shown in Figure 11, the use of 900MHz radio results in a better throughput than when 2.4GHz radio is employed, even for a longer distance. There is a larger variation when using 900MHz radio, which might be due to MAC-layer retransmission. The throughput of Ethernet-over-power line communication is very stable and typically maintained at about 600Kps. Using Wireless Mesh Network for Traffic Control Figure 9. Effect of number of hops on consecutive packet loss Figure 10. Effect of distance on consecutive packet loss Figure 11. Comparison of throughput for different technologies deploymenT issues In this section, we discuss some issues in terms of the deployment and maintenance of our testbed in an urban environment. deployment • Hardware. Many antenna connectors were held on by weak glue or crimp. Gradual stress (e.g. vibration) could eventually loosen the plug and degrade the signal before it is transmitted into the air. Some protection of the antenna plug might be necessary be necessary for an operational network to ensure there is no signal leakage from the antenna connector. In addition, while the appearance of the hardware might look identical, it is safer to check if the hardware does comply with the specification before starting using it. For example, some of Senao wireless cards used do not output a transmission power of 200mW as they should be according to the specification. 343 Using Wireless Mesh Network for Traffic Control • • Software. Most of the wireless measurements are based on readings from the wireless cards. However, while the hardware can be identical, different firmwares and drivers could introduce inaccuracy in the measurement results. If possible, one should try to validate the readings from a wireless card against the results from a spectrum analyser. Antenna locations. As described in Section 5, each node is equipped with three antennas, including two 2.4GHz (or one 2.4GHz and one 900MHz) omnidirectional antennas and one 3.5GHz directional antenna. To facilitate the ease of mounting, all three antennas are mounted on a pole and then mount this pole on the traffic light. Specifically, one omni-directional antenna is pointing upward and the other is pointing downward while the directional antenna is mounted in between. The location of antenna can have an effect on the link performance. Figure 12 shows the round-trip delays from node m2 to its neighboring node m3 using the omnidirectional antennas. The use of the lower antenna results in a higher loss and a larger variation. One possible explanation is that the upper antenna might be less obstructed and hence has a better connectivity. At 2.4GHz, a quarter wavelength is approximately 30cm. Antenna position changes in the range 10-30 cm can cause dramatic changes in signal strength, presumably due to the presence of standing waves in the vicinity of the traffic light pole or more specifically in the vicinity of metal stop signs and the like! If multiple antennas are deployed, it is essential to have a means for independently adjusting their position. maintenance • • Figure 12. Effect of antenna locations on round-trip delay 344 Remote management: Remote management is challenging, the following ways are provided to allow the user to access the nodes. To access the Linux-based Ethernet router: The router can be connected via the Unwired link over OpenVPN. In the case when the OpenVPN connection can not be established, given that a public IP address is obtained for each router from Unwired, one can connect to the router via its public IP address, although this could introduce a Using Wireless Mesh Network for Traffic Control • dependency on a dynamic DNS lookup. In addition, one can connect to the mesh node first and then connect to the router via the Ethernet or USB-to-serial link between the router and the motherboard. To access the mesh node: One can connect to the mesh node (i.e. the motherboard) by first connecting to the router and then connect to the motherboard via the Ethernet or USB-to-serial link. Being able to access the motherboard via its serial port is important since the Ethernet link might fail for various reasons. In addition, one can access the motherboard via its 802.11 interfaces from any reachable neighboring nodes. In addition, the following mechanisms are implemented to recover the system from failure. First, and the default way to recover from a failure is to login to the offending router or motherboard using one of the above methods, analyse the problem and/or reboot the node. Second, the watchdog timer support on the MB720 is used. In addition, Grub is setup to fall back to a stable backup image which is installed in a separate partition in case when the default image fails. Finally, the BIOS is configured to give top priority to PXE network boot but DHCP server is configured in a way that it does not provide PXE boot information in the default case. Therefore the node defaults to its second priority, which is to boot from the USB flash drive. However, in the event of a node failure (for example, due to a bad image) the DHCP server can be quickly reconfigured to support the PXE boot. Having rebooted the node using PXE, a new working disk image can be distributed to the node via frisbee or FTP. In practice, one might use FTP instead of frisbee since that frisbee introduce more control traffic overhead on the Unwired links, which charges for every bits send to the nodes. • Security: Security is a major concern especially when the wireless mesh testbed is sharing public spectrum with an average of 50+ external APs at each intersection. Furthermore, the testbed is effectively connected to the Internet via Unwired network, and exposed to various password attacks. In a live deployment for traffic control, the mesh security should be integrated with the traffic control system security model, which may include e.g. segmentation to contain the damage of a denial of service or break-in attack, combined with multiple levels of fallback to local control. conclusion In this chapter, we discuss some issues in deploying a testbed as a first step towards creating a fully functional wireless mesh network-based traffic control system. In addition, we describe some initial results of link characteristics of different technologies used on our testbed. Many research challenges such as latency, reliability, security and scalability are need to be addressed. reFerences Aarnet - australia’s research and education network. (2009). Retrieved from http://www.aarnet. edu.au/ Aguayo, D., Bicket, J., Biswas, S., Judd, G., & Morris, R. (2004). Link-level measurements from an 802.11b. Bicket, J., Aguayo, D., Biswas, S., & Morris, R. (2005). Architecture and evaluation of an unplanned 802.11bmesh network. In Proceedings of the 11th annual international conference on Mobile computing and networking(MOBICOM), Cologne, Germany. Available from http://www. pdos.lcs.mit.edu/papers/roofnet: mobicom05/ roofnet-mobicom05.pdf 345 Using Wireless Mesh Network for Traffic Control Camp, J., Robinson, J., Steger, S., & Knightly, E. (2006). Measurement driven deployment of a two-tier urban mesh access network. In Proceedings of ACM MobiSys 2006, Uppsala, Sweden. Available from http://networks.rice.edu/papers/ sys7122-camp.pdf Chakravorty, R., & Pratt, I. (2002). Performance issueswith general packet radio service. Journal of Communications and Networks (JCN), 4(2). Available from http://www.cl.cam.ac.uk/Research/ SRG/netos/papers/comob-web/2002-jcn.pdf Mobile taiwan applications promotion project m-taiwan. Retrieved from http://www.pwlan.org. tw/mp.asp?mp=3 Rapport, T. S. (1996). Wireless communications principles and practice. Upper Saddle River, NJ: Prentice Hall. Damn small linux not (dsl-n), (n.d.). Retrieved from http://www.damnsmalllinux.org/dsl-n/ Raychaudhuri, D., Seskar, I., Ott, M., Ganu, S., Ramachandran, K., Kremo, H., et al. (2005). Overview of the orbit radio grid testbed for evaluation of next-generation wireless network protocols. In Proceedings of the IEEE Wireless Communications and Networking Conference, New Orleans, LA, USA. Ganguly, S., Navda, V., Kim, K., Kashyap, K., Niculescu, D., Izmailov, R., et al. (2006). Performance optimizations for deploying voip services in mesh networks. Retrieved from http://www.wings. cs.sunysb.edu/%7Eanand/papers/jsac06.pdf Singh, M., Ott, M., Seskar, I., & Kamat, P. (2005). Orbit measurements framework and library (oml): Motivations, design,implementation, and features. In Proceedings of IEEE Tridentcom 2005, Trento, Italy. Jardosh, A., Ramachandran, K., Almeroth, K. C., & Belding-Royer, E. M. (2005). Understanding congestion in ieee 802.11b wireless networks. In Proceeding of ACM SIGCOMM Internet Measurement Conference, Berkeley, CA. Retrieved from http://moment.cs.ucsb.edu/amitj/jardoshimc2005.pdf Tropos networks. (2009). Retrieved from http:// www.tropos.com Locust world. (2009). Retrieved from http://www. locustworld.com Lundgren, H., Ramachandran, K., Belding-Royer, E., Almeroth, K., Benny, M., Hewatt, A., Touma, A., & Jardosh, (2006). Experiences from building and using the ucsb meshnet testbed. IEEE Wireless network. Available from http://user.it.uu.se/ [UNKNOWN ENTITY &tildent;]henrikl/publications.html mesh network. 346 Unwired wireless. (2009). Retrieved from http:// www.unwired.com.au/ Weber, S., Cahill, V., Clarke, S., & Haahr, M. (2003). Wireless ad hoc network for dublin: A large-scale ad hoc network test-bed. ERCIM News, 54. Retrieved from http://www.ercim.org/ publication/ErcimNews/enw54/weber.html Section 7 Mobility Model, Simulation, and Security 348 Chapter 20 Mobility Models of Vehicular Networks Kun-Chan Lan National Cheng Kung University, Tainan, Taiwan, R.O.C. absTracT A key component for VANET simulations is a realistic vehicular mobility model that ensures conclusions drawn from simulation experiments will carry through to real deployments. However, VANET simulations raise many new questions about suitable levels of details in simulation models. To get accurate results, the mobility models of Vehicular Networks should be as realistic as possible, and involve road-maps with all constraints and facilities related to the vehicular movement. In this chapter, the authors provide an overview of some mobility models that are relevant to VANETs. The criteria of applicability they consider here is the employment of road maps, and thus limiting the nodes movements into the routes, instead of moving them in a wide open area. They compare different models based on the parameters they use. For instance, some models use traffic control mechanisms (stop signs or traffic lights) at route intersections, and some just assume continuous movement at these points. Some assume routes to be single-lane, some others support multi-lanes routes. Some define the security distance, while others just ignore this parameter. inTroducTion To get accurate results, the model should be as realistic as possible, and involve road-maps with all constraints and facilities related to the vehicular movement. In this chapter, we provide an overview of some mobility models that are relevant to DOI: 10.4018/978-1-60566-840-6.ch020 VANETs. One of the realistic applications of mobile Ad hoc Networking (MANET) is Vehicular Ad hoc Networking (VANET). In VANET, communications between nodes do not rely on any dedicated infrastructure. Although it is self-organized and easy to deploy, the infrastructureless vehicular network introduces many challenges that should be tackled before real implementations. For instance, to allow communications between nodes (vehicles) which are Copyright © 2010, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global is prohibited. Mobility Models of Vehicular Networks out of the power range of each other, some other intermediaries should act as routers to remedy the lack of dedicated routers. Thus, a distributed routing protocol needs to be employed. It is mandatory before passing to the real deployment of a routing protocol as well as any other protocol or application) to evaluate it by simulation. The faithfulness of the simulation results depends on the realism of the parameters and the models used in the simulation, particularly on the mobility model. This can be defined as the pattern that establish the nodes movement within simulation area during a simulation. The earlier mobility models used in MANET simulation assume the terrain to be without obstacles, and nodes to be able to move freely in the whole rectangular stimulation area. For example, Random way-point (RWP) (David et al., 1996) is a typical example of such a kind of models, which is largely used in the literature, and available in many network simulators (ns2, GloMoSim, etc.). This model defines the pausetime parameter, so that each node has phases of movement and others of pause. At the beginning, the node selects randomly and uniformly a destination where it goes using a random and constant speed, which is also selected for each movement following a uniform distribution. Once it reaches this destination it stays there for the pause-time duration, then repeats the process. It has been illustrated that this model engenders, after a given simulation time, a spatial distribution of nodes concentrated around the center of the simulation area (Bettstetter et al., 2001 ; Bettstetter et al., 2003). Generally speaking, the assumption of an open terrain is realistic for some applications of pedestrians, but it is inappropriate for VANET. More recent studies propose new models constrained to routes and obstacles, and thus are more suitable for VANET. In (Camp et al., 2002) Camp et al. discuss a variety of mobility models used to evaluate ad hoc networks, and split them up into two categories: entity models and group models. The authors show by simulation how the choice of a mobility model can have a significant effect on the performance results. However, except city section (Zheng et al., 2004), all the other models presented in this survey are inappropriate for VANETs. In (Bettstetter et al., 2001; Zheng et al., 2004) the authors classify the models according to the randomness of speeds and directions, and divide them into i) trace-based (deterministic) models, ii) constrained topology based models and iii) statistical (fully random) models. Based on this classification, (Zheng et al., 2004) provides a more recent survey of MANET’s mobility models, presenting some ones that consider obstacles in the simulation area. These models are suitable to simulate pedestrian movements, but not vehicles. In (Boudec et al., 2005), the authors provide a mathematical and simulation investigations into some of the statistical (random) models, and combine them in the so-called random trip. The most recent research works on mobility models focus on the vehicular ad hoc network application of MANET. (Luo et al., 2004) provides a general survey of VANET, and the existing challenges to overcome before the real deployment. In (Chisalita et al., 2004) the authors discuss the usefulness of VANET to ensure the vehicular traffic safety and facilities, as well as the advantages it provides compared to the other centralized technologies. Some specified applications have been proposed, such as the discovery of free parking places (Caliskan et al., 2006). Regarding the mobility models, some new ones have been specifically proposed for VANET, such as (Amit Kumar Saha et al., 2004 ; Choffnes et al., 2006 ; Mahajan et al., 2006; Gorgorin et al., 2006; Karnadi et al., 2007). In this manuscript, we first present and discuss these novel models, as well as those proposed for the general MANET that apply to VANET. We also present a new vehicular traffic simulator we implemented to generate movement trace files usable by some well-known network simulator, notably GloMoSim (Zeng et al., 1998) and ns2 (The Network Simulator, 2009), then we use this tool along with GloMoSim to conduct a simulation comparative study. Our main contribution here is 349 Mobility Models of Vehicular Networks the investigation into the overtaking impact on the network performance. In the following we provide an overview on the models that were either specially devoted to VANETs, or proposed in the general MANET context but usable in VANET. The criteria of applicability we consider here is the employment of road maps, and thus limiting the nodes movements into the routes, instead of moving them in a wide open area. As we will see, the considered parameters differ from one model to another. For instance, some models use traffic control mechanisms (stop signs or traffic lights) at route intersections, and some just assume continuous movement at these points. Some assume routes to be single-lane, some others support multi-lanes routes. Some define the security distance, while others just ignore this parameter. Freeway is a generated-map-based model, defined in (Bai et al., 2003). The simulation area, represented by a generated map, includes many freeways, each side of which is composed of many lanes. No urban routes, thus no intersections are considered in this model. At the beginning of the simulation, the nodes are randomly placed in the lanes, and move using history-based speeds. The speed of each vehicle smoothly changes following a random acceleration. In addition to the realism related to the acceleration and the history based speed, the model defines a security distance that should be maintained between two subsequent vehicles in a lane. If the distance between two vehicles is less than this required minimal distance, the second one decelerates (its acceleration is forced to be negative) and let the forward vehicle moves away. The change of lanes is not allowed in this model. The vehicle moves in the lane it is placed in until reaching the simulation area limit, then it is placed again randomly in another position and it repeats the process. This scenario is definitely unrealistic. 350 manhaTTan This is also a generated-map-based model, introduced in (Bai et al., 2003) to simulate an urban environment. Before starting a simulation, a map containing vertical and horizontal roads is generated. Each of these latter includes two lanes, allowing the movement in two directions (north/south for the vertical roads and east/ west for the horizontal ones). At the beginning of a simulation, vehicles are randomly put on the roads. Afterwards, they move continuously according to history-based speeds (exactly like Freeway). When reaching a crossroads, the vehicle randomly chooses a direction to follow. That is, continuing straightforward, turning left, or turning right. The probability of each decision is set by the authors respectively to 0.5, 0.25, 0.25. The security distance is also used in this model, and nodes follow the same strategy as in the freeway model to keep this distance. But contrary to the previous model, a vehicle can change a lane at a crossroads. Nonetheless, there is no control mechanism at these points (crossroads), where nodes continue their movements without stopping, which is unrealistic. ciTy secTion mobiliTy (csm) CSM (Zheng et al., 2004) can be viewed as a hybrid model between RWP and Manhattan, as it introduces the principle of RWP, especially the pause-time and random selection destination, within a generated-map-based area. At each step of the vehicle’s movement a random point is selected from the generated road map, toward which it moves following the shortest path. After reaching that destination, it remains there for a pause-time, then repeats the process. The speeds of nodes are constrained by the security distance, along with the maximum speed limit of the road. Mobility Models of Vehicular Networks real map model (rmm) Thus far, we have presented models based on virtual generated maps. RMM (Amit Kumar Saha et al., 2004) is very similar to CSM, but indeed it uses real maps, obtained from the TIGER/Lines database (Amit Kumar Saha et al., 2004). For each route segment 1, the coordinates are extracted and converted using the Mercator projection (Amit Kumar Saha et al., 2004). The extracted points are then presented by a graph, where the crossroads are presented by vertices, and routes by weighted arcs. The weight of each arc is dynamically calculated in such a way to mimic the estimated time required for a vehicle to move over the corresponding segment, which is proportional to its maximum authorized speed, its distance, and the number of vehicles it currently contains. Therefore, the lower the weight, more the vehicles move freely in the segment. Note that the maximum authorized speed of a route segment depends on its type. Finally, we mention that like all the previous models, RUM defines no control mechanisms at crossroads. sTop sign model (ssm) Contrary to the previous models, SSM (Mahajan et al., 2006) integrates a traffic control mechanism. In every crossroads, a stop signal is put, which obliges vehicles to slow down and make a pause there. This model is based on real maps of the TIGER/Lines database, but all roads are assigned a single lane in each direction. A vehicle should never overtake its successor (like in all the models presented before), and should tune its speed to keep the security distance. If many vehicles arrive at an intersection at the same time, they make a queue, and each one waits for its successor to traverse the crossroads. This results in gathering of nodes, and hugely affects the network connectivity as well as the vehicle mobility (average speeds). According to the authors (Mahajan et al., 2006), the problem with this model is the unrealistic disposition of the stop signals, since it is impossible to find a region with stop signals at each intersection. Therefore, they proposed TSM (Mahajan et al., 2006), which we describe hereafter. TraFFic sign model (Tsm) In this model, stop signals are replaced by traffic lights. A vehicle stops at a crossroads if it encounters a red stoplight, otherwise it continues its movement. When the first vehicle reaches the intersection, the light is randomly turned red, with probability p (thus turned green with probability 1 − p). If it turns red, it remains so for a random delay (pause-time), forcing the vehicle to stop as well as the ones behind it. After the delay, it turns green then the nodes traverse the crossroads one after the other until the queue is empty. When the next vehicle arrives at the crossroads the process is repeated. TSM and SSM have been evaluated by simulation with ns2 (The Network Simulator, 2009). The results show that the two models are not significantly influenced by the speed of nodes (maximum speeds). This is due to the traffic control models, which slow down the nodes and give more stability to the network (Mahajan et al., 2006). When increasing the pause-time at the intersections, the authors remarked that the performances improved for both models, and that SSM gives better results than TSM when using the same pause-time. The authors argue this by the fact that in SSM nodes always stop at the intersections, unlike TSM. Nevertheless, in reality the pause-time for stop signals is shorter than that of traffic lights, which makes TSM more stable indeed (Mahajan et al., 2006). sTraw TraFFic STRAW (Choffnes et al., 2006) is also a model relying on real maps of TIGER/Line. Like the 351 Mobility Models of Vehicular Networks other models (except freeway), the roads include one lane in each direction, and is divided into segments. The model is basically composed of three modules: intra-segment mobility manager, inter-segment mobility manager, and finally the route management and execution module. At the beginning of the simulation, the nodes are placed randomly one behind the other. Then they move using the car following model (Choffnes et al., 2006) such that they try to accelerate until reaching the maximum speed of the segment. The first module manages this movement until reaching an intersection. The security distance is maintained, and the overtaking is not allowed. At crossroads the vehicles always slow down, even when they change a segment and turn without a full stop, which is realistic. The second module defines the traffic control mechanism, which includes both stop signals and traffic lights, put on crossroads according to the class of the intersected routes. In addition to this usual control form, the module makes sure that the next segment to take contains enough available space before moving the vehicle toward it. If it is fully busy, the vehicle waits at the crossroads (at the end of the first segment). The last module selects the routes to be taken by each vehicle during the simulation. It implements two approaches: simple straw and straw OD. In the first one, the direction is randomly selected at each intersection. That is, when reaching an intersection, the vehicle randomly decides whether to continue straightforward, or to turn and change the route. On the other hand, in the second approach a destination is selected toward which the vehicle moves using the shortest path. The simulation study made by the authors (Choffnes et al., 2006) show that when using STRAW the reception ratio decreases from 43% up to 53% compared to movements in an open area. The results of this simulation also illustrate that the routes arrangement has an impact. Scenarios with a high number of crossroads slow down the average speeds of nodes, which improves the reception ratio. 352 moVe MOVE (Karnadi et al., 2007) is a VANET’s mobility model that uses as the compiler SUMO (SUMO Simulation of Urban Mobility, 2009), which is a realistic vehicular traffic simulation model. SUMO is an open-source application implemented with java, that integrates many realistic parameters, such as realistic accelerations, the usage of real maps reflecting several types of routes (with multiple lanes), as well as traffic lights defining priorities between vehicles. Basically, MOVE is composed of two components: the road map editor and the vehicle movement editor. The former serves to manually and randomly generate a road map, either from TIGER/ line files or Google earth files, whereas the latter allows to specify the properties of each vehicle, like the maximum speed, the acceleration, the probability of turning at crossroads, the path to take etc. The information collected by the two editors is sent to the SUMO compiler, then a trace file in ns-2 or Qualnet format is generated. MOVE has been compared by simulation to RWP using AODV. The results show that MOVE causes a low reception rate. gorgorin eT al. model In addition to all the realistic parameters of the previous models, this one (Gorgorin et al., 2006) implements an overtaking mechanism within multi-lane segments. A vehicle always tries to move on the most right lane (the lowest rang), except in case of overtaking during which it moves left, and intersections in urban environments where it chooses the lane according to the next direction. A hierarchy of vehicle states is defined, respectively: free driving, approaching, following, and braking (in the order). When a vehicle is in another state than the free driving, it checks whether higher lanes allow it to pass to a higher state, and thus moves to the left lane to make an Mobility Models of Vehicular Networks Table 1. Feature\ Model Real maps # lanes/direction Freeway Manhattan CSM RUM SSM TSM STRAW MOVE Gorgorin no no no yes yes yes yes yes yes many one one one one one one many many Intersections no yes yes yes yes yes yes yes yes Changing lanes at intersections no yes yes yes yes yes yes yes yes Traffic control no no no no stop signs traffic lights both traffic lights+ priority both Overtaking no no no no no no no no yes Security distance yes Yes yes no yes yes yes yes yes Pause-time no No yes no yes yes yes yes yes overtaking. Identically, a vehicle in a state different than braking checks whether the right lane allows it to at least stay in the same state and then moves right. Moreover, the model allows to specify the driver type, which affect many parameters of the vehicles (speed, acceleration, etc.). Finally, note that the model includes both traffic lights and stop signals at intersections. One of these two different control mechanisms is put at each intersections according to the types of the intersecting segments. The most important parameter added in this model is the overtaking mechanism. However, no study investigating this issue has been done yet. Table 1 summarizes the features of all the models presented in this section. reFerences Amit Kumar Saha, D. B. J. (2004). Modeling mobility for vehicular ad-hoc networks. In the 1st ACM international workshop on Vehicular ad hoc networks (pp. 91–92). New York: ACM Press. Bai, F., Sadagopan, N., & Helmy, A. (2003). Important: a framework to systematically analyze the impact of mobility on performance of routing protocols for ad hoc networks. In The 22th IEEE Annual Joint Conference on Computer Communications and Networking INFOCOM’03, (pp. 825–835). Bettstetter, C. (2001). Smooth is better than sharp: a random mobility model for simulation of wireless networks. In Proceedings of the 4th ACM international workshop on Modeling, analysis and simulation of wireless and mobile systems (MSWIM ’01) (pp. 19–27). New York: ACM Press. Bettstetter, C., Resta, G., & Santi, P. (2003). The node distribution of the random waypoint mobility model for wireless ad hoc networks. IEEE Transactions on Mobile Computing, 2(3), 257–269. doi:10.1109/TMC.2003.1233531 Boudec, J.-Y. L., & Vojnovic, M. (2005). Perfect simulation and stationarity of a class of mobility models. In The 24th IEEE Annual Joint Conference on Computer Communications and Networking INFOCOM’05, (pp. 2743–2754). 353 Mobility Models of Vehicular Networks Caliskan, M., Graupner, D., & Mauve, M. (2006). Decentralized discovery of free parking places. In Proceedings of the 3rd International ACM Workshop on Vehicular ad hoc networks, VANET’06, (pp. 30–39). New York: ACM Press. Camp, T., Boleng, J., & Davies, V. (2002). A survey of mobility models for ad hoc network research. Wireless Communications and Mobile Computing (WCMC): Special issue on Mobile Ad Hoc Networking: Research . Trends and Applications, 2(5), 483–502. Chisalita, I., & Shahmehri, E. (2004). Vehicular communication: A candidate technology for traffic safety. In IEEE International Conference on Systems, Man and Cybernetics, (pp. 3903–3908). Choffnes, D. R., & Bustamante, F. E. (2006). An integrated mobility and traffic model for vehicular wireless networks. In The 2nd ACM international workshop on Vehicular ad hoc networks, VANET’05, (pp. 69–78). New York: ACM Press. David, B., & David, A. (1996). Dynamic source routing in ad hoc wireless networks. In Mobile Computing, (vol. 35, pp. 153–181). Kluwer Academic. Davies, V. (2000). Evaluating mobility models within an ad hoc network. Master’s thesis, Colorado School of Mines, Colorado, USA, Tech. Rep. Gorgorin, C., Gradinescu, V., Diaconescu, R., Cristea, V., & Ifode, L. (2006). An integrated vehicular and network simulator for vehicular ad-hoc networks. In the 20th European Simulation and Modelling Conference. 354 Johansson, P., Larsson, T., & Hedman, N. (1999). Scenario-based performance analysis of routing protocols for mobile ad-hoc networks. In MOBICOM 99, The Fifth Annual ACM/IEEE International Conference on Mobile Computing and Networking, Seattle, WA (pp. 15–19). Karnadi, F. K., Mo, Z. H., & Lan, K. C. (2007). Rapid generation of realistic mobility models for vanet. In IEEE Wireless Communications and Networking Conference. Luo, J., & Hubaux, J.-P. (2004). A survey of intervehicle communication. School of computer and Communication Sciences, EPEL, Tech. Rep. IC/2004/24. Mahajan, A., Potnis, N., Gopalan, K., & Wang, A.I. A. (2006). Urban mobility models for vanets. In Proceedings of the 2nd IEEE International Workshop on Next Generation Wireless Networks. SUMO Simulation of Urban Mobility. (2009). Retrieved from http://sumo.sourceforge.net/ The Network Simulator. ns2.(2009). Retrieved from http://www.isi.edu/nsnam/ns/ Zeng, X., Bagrodia, R., & Gerla, M. (1998). Glomosim: A library for the parallel simulation of large-scale wireless networks. In The 12th Workshop on Parallel and distributed Simulation. PADS’98, Banff, Alberta, Canada (pp. 154–161). Zheng, Q., Hong, X., & Ray, S. (2004). Recent advances in mobility modeling for mobile ad hoc network research. In Proceedings of the 42nd ACM annual Southeast regional conference (ACM-SE), (70-75). New York: ACM Press. 355 Chapter 21 MOVE: A Practical Simulator for Mobility Model in VANET Kun-Chan Lan National Cheng Kung University, Tainan, Taiwan, R.O.C. absTracT Vehicular Ad-Hoc Network (VANET) is surging in popularity, in which vehicles constitute the mobile nodes in the network. Due to the prohibitive cost of deploying and implementing such a system in real world, most research in VANET relies on simulations for evaluation. A key component for VANET simulations is a realistic vehicular mobility model that ensures conclusions drawn from simulation experiments will carry through to real deployments. However, VANET simulations raise many new questions about suitable levels of details in simulation models for nodes mobility. In VANET simulations, the mobility models used affect strongly the simulation output. The researchers need to decide what level of details are required for their simulations. In this chapter, the authors introduce a tool MOVE that allows users to rapidly generate realistic mobility models for VANET simulations. MOVE is built on top of an open source micro-traffic simulator SUMO. The output of MOVE is a realistic mobility model and can be immediately used by popular network simulators such as ns-2 and Qualnet. The authors show that the simulation results obtained when using a realistic mobility model such as MOVE are significantly different from results based on the commonly used random waypoint model. In addition, the authors evaluate the effects of details of mobility models in three case studies of VANET simulations (specifically, the existence of traffic lights, driver route choice and car overtaking behavior) and show that selecting sufficient level of details in the simulation is critical for VANET protocol design. DOI: 10.4018/978-1-60566-840-6.ch021 Copyright © 2010, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global is prohibited. MOVE inTroducTion VEHICULAR Ad-Hoc Network (VANET) communication has recently become an increasingly popular research topic in the area of wireless networking as well as the automotive industries. While it is crucial to test and evaluate protocol implementations in a real world environment, simulations are still commonly used as a first step in the protocol development for VANET research. Several communication networking simulation tools already existed to provide a platform to test and evaluate network protocols, such as ns-2 (T. N. S. ns 2, 2009), OPNET (O. Simulator, 2009) and Qualnet (Q. N. Simulator, 2009). However, these tools are designed to provide generic simulation scenarios without being particularly tailored for applications in the transportation environment. On the other hand, in the transportation arena, simulations have also played an important role. A variety of simulation tools such as PARAMICS (P. M. T. Simulation, 2009), CORSIM (CORSIM, 2009) and VISSIM (P. simulation VISSIM, 2009) etc have been developed to analyze transportation scenarios at the micro- and macro-scale levels. However, there was little effort in integrating communication techniques and scenarios in a realistic transportation simulation environment. One of the most important parameters in simulating ad-hoc networks is the node mobility. It is important to use a realistic mobility model so that results from the simulation correctly reflect the real-world performance of a VANET. A realistic mobility model should consist of a realistic topological map which reflects different densities of roads and different categories of streets with various speed limits. Another important parameter should be modeled is the obstacles. In the real world, a vehicle node is typically constrained to streets which are separated by buildings, trees or other objects. Such obstructions often increase the average distance between nodes as compared to that in an open-field environment. In addition, each vehicle needs to decide a turning directions 356 at the intersection (e.g. turn left, turn right or go straight). Such a turning model could have an effect on the congestion of the road as well as on the clustering of the vehicles. Furthermore, a smooth deceleration and acceleration model should be considered since vehicles do not abruptly break and move. Some prior studies (Saha et al., 2004; Heidemann et al., 2001) have shown that a realistic model is critical for accurate network simulation results. Selecting appropriate level of details in the mobility model for a VANET simulation is a critical decision. Unrealistic mobility model can produce simulations that are misleading or incorrect. On the other hand, adding details requires time to implement and debug. In addition, it might increase simulation complexity, slow down simulation, and distract the research problem at hand. In this chapter, we develop a tool MOVE (MObility model generator for VEhicular networks) to facilitate users to rapidly generate realistic mobility models for VANET simulations. MOVE provides an environment that allows the user to quickly pinpoint incorrect details and manage details overhead. Our tool is built on top of an open source micro-traffic simulator SUMO (S. S. of Urban Mobility, 2009). The output of MOVE is a mobility trace file that contains information of realistic vehicle movements which can be immediately used by popular simulation tools such as ns-2 or qualnet. MOVE allows users to rapidly generate realistic VANET mobility models in two aspects:- by interfacing with real world map databases such as TIGER (Topologically Integrated GEographic Encoding and Referencing) (TIGER, 2009) and Google Earth, MOVE allows the user to conveniently incorporate realistic road maps into the simulation. In addition, by providing a set of Graphical User Interfaces that automate the simulation script generation, MOVE allows the user to quickly generate realistic simulation scenarios without the hassle of writing simulation scripts as well as learning about the internal details MOVE Figure 1. The architecture of MOVE of the simulator. The architecture of MOVE is shown in Figure 1. In this chapter, we first describe the architecture implementation of MOVE (Section 2). We then compare MOVE against the commonly used random waypoint model and show that a realistic mobility model is critical for VANET simulations. We present three case studies that consider three different scenarios including the existence of traffic light, driver route choice at the intersection, and car overtaking behavior. We discuss how these details affect the network topology and resultingly the performance of VANET in the simulation (Section 3). trips of vehicles and the route that each vehicle will take for one particular trip. We currently support three different methods to define the vehicle movements– the vehicle movement patterns can be manually created by the user, generated automatically, or specified based on a bus time table to simulate the movements of public transport. The information users input in the Map Editor and the Vehicle Movement Editor is then fed into SUMO to generate a mobility trace which can be immediately used by a simulation tool such as ns-2 or qualnet to simulate realistic vehicle movements. Users can also visualize the generated mobility trace by clicking on the “Visualization” button on the main menu, as shown in Figure 3. This is useful for observing the details of vehicle movement (e.g. cars’ overtaking behavior) and debugging (e.g. to avoid incorrect setting of simulation parameters). Figure 2. Mobility model generator archiTecTure MOVE is currently implemented in Java and runs atop an open-source micro-traffic simulator SUMO. MOVE consists of two main components: the Map Editor and the Vehicle Movement Editor, as shown in Figure 2. The Map Editor is used to create the road topology. Currently our implementation provides three different ways to create the road map – the map can be manually created by the user, generated automatically, or imported from existing real world maps such as publicly available TIGER database from U.S. Census Bureau (TIGER, 2009). The Vehicle Movement Editor allows the user to specify the 357 MOVE Figure 3. Visualization of vehicle movements Finally, one of the major overhead before one can start conducting research using simulations is to learn about the internal details of the simulator and write customized simulator-specific scripts to generate various simulation scenarios for the research problem under study (Floyd et al., 2001). To reduce such an overhead, MOVE provides an interface to automatically generate simulation scripts on the fly based on the parameters that the user inputs into MOVE. We currently support auto-generation of ns-2 and qualnet simulation scripts, as shown in Figure 4. Figure 4. Traffic model generation 358 map editor In MOVE, the road map can be generated manually, automatically or imported from a real world map. Manual generation of the map requires inputs of two types of information, nodes and edges. A ”node” is one particular point on the map which can be either a junction or the dead end of the roads. Furthermore, the junction nodes can be either normal road junctions or traffic lights. The edge is the road that connects two points (nodes) on a map. The attributes associated with an edge include speed limit, number of lanes, the road priority and the road length. Figure 5 shows snapshots of nodes editor and edge editor. We have also integrated Google Earth into MOVE to facilitate the creation of nodes in a realistic setting. Google Earth is a tool that enables its user to view the satellite image map of any place on earth. One of the functionality that Google Earth provides is called “placemark” which allows the user to put a mark on any location of the Google Earth map. Each placemark contains the longitude and latitude information of the selected locations and can be saved into a file in KML format (K. tutorial). Hence, one can define the node location on the Google map and then extract the node information by processing the saved KML file. This allows MOVE users to generate a map for MOVE Figure 5. Road map generation any real-world road on earth for their simulations. Figure 6 shows an example of using Google Earth to generate nodes for the major intersections in the Eastern Suburb of Sydney.Figure 5. Road map generation The road map can also be generated automatically without any user input. Three types of random maps are currently available: grid, spider, and random networks. There are some parameters associated with different types of random maps such as number of grids and the number of spider arms and circles. Finally, one can also generate a realistic map by importing real world maps from publicly available database. We currently support the TIGER maps which are available from U.S. Census Bureau. Figure 7 shows a grid map generated from the random map generator and a street map in the Houston area based on a TIGER database file. inter-departure time of the vehicles originating from the starting road. In addition, a MOVE user can define the probability of turning to different directions at each junction (e.g. 0.5 to turn left, 0.3 to turn right and 0.2 to go straight) in the editor. Figure 8(a) shows a snapshot of the Flow definition Editor. One can also generate vehicle movement manually using the Vehicle Movement Editor which allows users to specify several properties of vehicle routes including the number of vehicles in a particular route, vehicle departure time, origin and destination of the vehicle, duration of the trip, Figure 6. Generating realistic map using Google Earth Vehicular movement editor The movements of vehicles can be generated automatically or manually using the Vehicle Movement Editor. To generate vehicle movement automatically, one needs to first define a vehicle flow which describes a fleet of vehicles toward the same direction. The parameters of each flow consist of the starting road and destination of the vehicle fleet, the time to start and end the vehicle flow, the number of vehicles in the flow and the 359 MOVE Figure 7. Road map generation using the Map Editor vehicle speed (including acceleration, deceleration and maximum speed), etc. Figure 8(b) shows a snapshot of the Vehicle Movement Editor. Note that, in addition to simulating vehicle-to-vehicle communication, our tool is also useful for simulations of vehicleto-infrastructure communication (e.g. the communication between mobile nodes and road-side static gateway nodes). A static node can be created in MOVE by assigning the vehicle with a maximum speed of zero in the Vehicle Movement Editor. On-board communication has recently become an increasingly popular research topic. A new paradigm of Networks in Motion is quickly attracting interest from the research community Figure 8. Vehicle movement generation 360 and is also being viewed as a viable commercial solution for extending Internet services to public transport passengers. MOVE allows users to enter the bus time table to simulate the movements of public transport. We model buses as a group of vehicles which have similar parameters such as speeds, routes, etc, associated with it as other vehicles. In addition, to model the bus time table, one should define the departure times of the first and the last bus and the bus inter-arrival time (which is assumed to be constant in our implementation). Figure 9 shows the editor for entering the bus route information. MOVE goal is to inject the simulation with as much detail as possible to provide a ”realistic” mobility model. We find mobility MOVE Figure 9. Bus route generation model acts a leading role in VANET, it determine the nodes location, density, and direction etcetera that influence directly VANET performance. Yet a ”fully realistic” simulation is not possible, in mobility model aspect due to human behavior is hardly description (i.e. mood, sex, age, etc.). In addition, there have a lot of accident in the real world, it also changed mobility model. Therefore, simulation designers must restrict the level of detail in some place. The challenge is to recognize what level of detail does not influence answers to the design questions at hand. eValuaTion In this section, we evaluate the impact of mobility models generated by MOVE on the performance of ad-hoc routing protocol. First, we compare the performance of AODV (Perkins et al., 2003) when used with the random waypoint model to that using the MOVE mobility model. The simulation experiments were carried out in ns-2. Each simulation lasts for 900 seconds. We generated scenarios for 150 nodes moving in an area of 4 square kilometres. We varied the number of source nodes from 10 to 50, each of which is a CBR traffic source transmitting UDP Figure 10. Comparison between Random Waypoint and model generated by MOVE packets of a size 64 bytes at the rate of 4 packets per second. All nodes use 802.11 MAC operating at 2Mbps. The transmission range is 250m. The propagation model employed in the simulation is the log normal shadowing model. We used a path loss exponent 2.56 with standard deviation 4.0 based on real-world measurement data from an inter-vehicle experiment we previously carried out in Sydney suburban area. The road topology generated by MOVE is based on the TIGER database data. Figure 10 shows the packet delivery ratio of AODV with different number of traffic sources. Each data point represents the average of six runs and the error bars represent the range of observed packet delivery ratios. Overall, the packet delivery ratios increase as the number of traffic sources increases, which suggest a higher density of nodes can increase the network performance as long as the increasing density does not create more radio interference. In addition, the packet delivery ratios of AODV based on MOVE mobility models are lower than when the Random Waypoint model is used. The results generated from MOVE also have larger variations. The larger variance in MOVE data points is possibly due to unstable network connectivity imposed by constrained node movements by roads and traffic control mechanisms (such as traffic lights). Figure 10 clearly shows 361 MOVE that the simulation results using a more realistic mobility model can be drastically different from that using a simplistic open field model. Note that our results are also consistent with prior work (Choffnes et al., 2005). Second, we evaluate the effects of details of mobility models in three case studies. Specifically, we set out to understand how the existence of traffic lights, driver route choice and car overtaking behavior affect the VANET simulation results. The number of nodes in our simulations is 300 and the simulation time lasts for 1000 seconds. The roads created in the simulation have two lanes. Traffic lights simulation In real world, traffic lights are used to regulate traffic flow moving in different directions. The existence of traffic lights tends to create a “clustering” effect. In other words, places where there is a traffic light are likely to have a higher node density due to that vehicles are forced to stop at the traffic light to wait for the light to turn green. Intuitively, a high node density might improve the network connectivity. On the other hand, a higher node density might also suggest a higher chance for packet collision since more nodes might be transmitting at the same time. In addition, the distance between two adjacent traffic lights can have a significant effect on the network connectivity. Specifically, the network can be “fragmented” by the traffic lights when the Figure 11. Clustering effect due to the traffic light 362 radio transmission range is smaller than the distance between two adjacent clusters. In other words, a link breakage can happen when the inter-cluster distance is larger than the radio coverage. Figure 11 shows the distribution of the number of neighboring nodes when ten traffic lights are included in the simulations. Our results show that each node has twice the number of neighboring nodes when traffic lights are simulated, as compared to the case when traffic lights are not simulated. Here we define a “neighboring node” as the node which is within the radio range of a vehicle. Having a larger number of neighboring nodes typically suggests a better network connectivity. As shown in Figure 12, the packet delivery ratio is improved when the traffic lights are simulated. Note that in this simulation the distance between Figure 12. Effect of traffic light Figure 13. Effect of inter-traffic-light distance MOVE two adjacent traffic lights is shorter than the given radio range. In addition, we observe that the number of packet collisions increase as we increase the number of traffic sources. As a result, the packet delivery ratio decrease when there are more traffic sources. To understand the effect of inter-cluster distance on the simulations results, we increase the distance between two adjacent traffic lights (from 200m to 400m) so that the inter-cluster distance is larger than the effective radio distance. As shown in Figure 13, in this scenario we observe frequent link breakage between two adjacent clusters which significant degrade the network performance. The effective radio range is around 250m in this experiment. overtaking simulation In real world, a driver normally has to decide his moving direction at an intersection. He can choose to either go straight, turn left, or turn right. MOVE allows a user to define the turning probability to different directions at each intersection (e.g. 0.5 to turn left,0.3 to go straight and 0.2 to turn left) in the Vehicle Movement Editor. As shown in Figure 14, we find that different choices of route directions can significantly change the simulation results (the x-y-z notation in Figure 14 means that the car has x% of chance to turn left, y% to go straight and z% to turn right). In real world, a faster vehicle can overtake some other slower ones when overtaking is allowed on a multi-lane road. Overtaking behavior can have a great effect on the network topology and should be considered. Specifically, when overtaking behavior is not allowed, it usually results in a chain-like topology and a shorter and uniform inter-vehicle distance (the uniform distance is due to that the vehicle needs to maintain a safe distance from the adjacent cars), which often suggests a better network connectivity. We observe a dramatic impact on the network performance when the overtaking behavior is allowed. In addition, we find that the effect of overtaking behavior is less significant when the network density is higher. As shown in Figure 15, the packet delivery ratios in overtaking-allowed scenario is close to results of no-overtaking scenario when we increase the number of nodes from 250 to 350. In summary, we show that details of mobility models such as the existence of traffic lights, driver route choice and car overtaking behavior can have a drastic impact on the VANET simulation results. We argue that the faithfulness of simulation results is proportional to the realism of the parameters and the models used in the simulations. Therefore, selecting appropriate level of details in the mobility model for a VANET simulation is a very important yet challenging task. Figure 14. Effect of driver route choice Figure 15. Effect of car overtaking behavior Turning simulation 363 MOVE relaTed work In this section, we have discussed mobile mobility model with level of different details and VAENT simulators. Mobility models play a critical role in VANET. This is because mobility affects routing and network performance etc. Therefore, the user should carefully choose suitable mobility model in applications. mobility model with level of different details Random Way-Point (RWP) (David et al., 1996) is the earlier mobility models used in MANET simulation. It consider without obstacles topography and nodes to be able to move freely in simulation area. RWP is widely to use mobility model in simulation (Das et al., 2000 ; Holland et al., 1999 ; Perkins et al., 1999,; Johansson et al., 1999.), which only suit mobile nodes in MANET. This is because VANET environment have obstacles and each vehicle restrict by street. Authors of (Hong et al., 1999) present a Reference Point Group Mobility (RPGM) model to depict the relationship among mobile hosts. This model based on Random Waypoint model to use checkpoint for movement. Although this model without any obstacle, but a suitable checkpoint scenario can model realistic mobility patterns. Unfortunate, a common and realistic mobility model found hard. (Bettstetter et al., 2001) present an improved Random Direction Model instead of Random Waypoint Model. It introduces a stop-turn-and-go behavior for vehicles or bicycles movement patterns that discuss alike vehicular nodes stop over at intersection. This model can extend to turning and stopping traffic light events in VANET. Similar to the previous literature, authors of (Royer et al., 2001) use simulation to present random direction model. Due to the simulations focus on mobile ad hoc network and without considering realistic map. In the other words, it not suit for VANET. In addition, Hong et al. (2001) discuss different mobility model as 364 follows: random walk (Zonoozi et al., 1997) (this model not discuss in the paper), RWP, RPGM, and their approach, mobility vector model. The mobility vector model’s advantage can use for different behavior, such as gravity model, location dependent model, targeting model, etc. Due to it have deceleration and change direction behaviors for more realistic mobility model. This meaning the model possesses basic characteristic for vehicular mobility. But this model still hard widely use in the real world. Although human drivers usually is intuitive, it yet need consider different traffic condition, such as traffic congestion or sparse. For instance, when the traffic condition is congestion, the vehicles only suit car following. Vice versa, overtaking most happen at traffic sparse. Therefore, this is important phenomena. Camp et al. (2002) survey variety of the mobility model in ad hoc networks, such as random walk mobility, RWP, random direction mobility, probabilistic random walk etc. This can sort two main models: entity models and group models for mobile node. Fortunately, VANET constrain by street that the mobility model become simple. But in the real world, the driver has indefinite behavior with sex, age, mood, habits etc. effects lead to different mobility pattern (e.g. overtaking, car following, etc.). The traffic light also affect vehicles movement in VANET. Bai et al. (2003 ; 2003, November) introduced RWP, group mobility, Freeway, and Manhattan mobility models. Different mobility models have especially application and characteristics as follows: group mobility use in military battlefield or motorcade of travel etc. Generally, this mobility model have cluster and both speed and direction. Freeway mobility model, this mobility model should have car-following and overtaking movement behaviors in general case, such that usually happen at traffic congestion or rarefaction. Nevertheless, the authors not mention car following and overtaking. The urban environment use Manhattan mobility model, the mobility model is a simple urban mobility model that the vehicles only move along the grid of horizontal MOVE and vertical streets on the map. Actually, the urban have a lot of traffic light to manage and control traffic flow, this issue discuss in evaluation section. Until now, we survey the mobility models that most use RWP and without considering obstacle. They are not suitable use on VANET. Saha et al. (2004) proposed based-on TIGER maps’ macro mobility model. It is only considering Dijkstra’s single source shortest path algorithm from source to destination location. Although use dynamic source routing protocol to evaluate performance by NS2. However, the scheme lacks traffic light mechanism and overtaking behavior. Jardosh et al. (2005) present an obstacle mobility model that the placement of obstacles that restrict movement and signal propagation, although it focuses discussion in mobility model of people, but the builds can instead of extracting data from TIGER files. This behavior spreads street restriction and build obstacles in VANET environment. It also use distance dynamic to adjust propagation models. This is for realistic simulation and good idea. (Stepanov et al., 2005) Using probabilities decide trips and use estimated travel time to choose movement path. This can increase realistic of mobility model that the probabilities can draw trips’ peak and off-peak in duration. In addition, the scheme hard get accurate travel time. This is because VANET fast changed. The travel time easy affect by traffic congestion or accident. In (Sven Jaap et al., 2005), the authors present city mobility model that build on IDM (Intelligent-Driver Model) (Treiber et al., 2000) with probability turning. It is meaning the model that support car following and turning. But it cannot provide network simulator traces, without considering traffic light and overtaking. Therefore, this model not enough call realistic mobility model. Zimmermann et al. (2005) proposed a voronoi-based mobility model for urban environments, this model is using a spatial area obtained movement paths which are computed by the multiple application of Voronoi graphs. Due to the vehicles movement is constrained to street by the voronoi graph, this phenomenon fit VANET network. In principle, this mobility model only show obstacle in VANET network. This result same as previous literature (Sven Jaap et al., 2005) not consider traffic light and overtaking. Street RAndom Waypoint (STRAW) (Choffnes et al., 2005), it is a random waypoint constraint by road. The scheme has considered traffic light control and car following model. It is also use shortest path algorithm to calculate movement path. The authors though have considered previous condition. But it still lack overtaking criterion that cause convey effect in street. However, it is not realistic and then only uses random method. Mahajan et al. (2006) proposed Stop Sign Model (SSM), Probabilistic Traffic Sign Model (PTSM), and Traffic Light Model (TLM). This paper focus in traffic control system. In SSM, each vehicle around the intersection must stop at the stop sign in duration time. PTSM use a probability p to deice the vehicle stop, this paper use with an empty street queue, but this model is not coordinating among different directions. In other words, PTSM can utilize different traffic condition to adjust traffic light time and decrease traffic congestion. The TSM similar to PTSM, it only changed at coordinating among different directions. Anyhow, any types’traffic control system both has pause time to control vehicle. This is important factor in traffic control system. When increasing the pause time at the intersections, each vehicle waiting time will elongate that increase traffic congestion. In the other street, this street’s vehicles not need stop over at intersection and move forward, this street almost no waiting time. This is a trade-off in different application. The authors’simulation results point out long pause-time which can increase packer delivery ratio. This only consider at street of local. This case is unrealistic. In (Potnis et al., 2006), the authors have simulated SSM and TSM mobility model to compare packet delivery ratio. On the other hand, (Marfia et al., 2007) Using CORSIM (CORSIM, 2009) and TRANSIMS (TRANSIMS) generate different mobility model in downtown, and then use Qualnet to compare network performance with vary transmission range etc. In addition, the authors also have discussed with infrastructure (i.e. AP) to improve packet delivery 365 MOVE ratio. Nevertheless, the context lack more detail of mobility model discussion, such as overtaking and lane change impact. VaneT simulators Groovesim (Mangharam et al., 2005) is a topography-accurate street-map based vehicle network simulator and is based on GrooveNet, a geographic routing protocol for vehicular networks. It provides several different modes of operation. In Drive Mode, GrooveSim can process data from a GPS unit to provide a real-time map of the vehicle’s current location. It can also be used as an emulator in Hybrid Simulation Mode where real vehicles on the road and virtual vehicles in the simulation can interact with each other. Groovesim also provides a tool for analyzing the simulation results. One limitation of Groovesim is that it is strongly tied to one specific routing protocol (i.e. GrooveNet), which limits its use for simulating other routing protocols in a VANET environment. In addition, GrooveSim does not provide mobility traces for network simulators. STRAW is an extension of SWANS (Scalable Wireless Ad Hoc Network Simulator) (Haas et al., 2005), a Java-based simulator for wireless simulations. STRAW contains simulation tools for generating mobility models and traffic models and is also able to use real street maps like TIGER data to build the road topology. However, currently the mobility models can be supported by STRAW is limited. For example, while STRAW supports multiple lanes, the vehicles are not allowed to change lane and the starting position is not configurable. Another drawback of this tool is its dependency on SWANS. Finally, STRAW does not provide any GUI that allows the users to visualize the movements of cars. BonnMotion (B. A. mobility scenario generation and analysis tool) is a simple tool that can be used to create and analyses mobility scenarios. Similar to MOVE, the mobility scenarios created by BonnMotion can be exported to ns-2 and qualnet. However, BonnMotion only models ba- 366 sic motion constraints and does not consider any micro-mobility. Furthermore, BonnMotion is a text-based application that runs on a command shell and does not provide any graphical user interfaces as MOVE does. Complementary to these previous efforts, our work emphasizes on creating a tool that allows users to rapidly generate realistic mobility models for VANET simulations. conclusion and FuTure work In this chapter, we describe a tool MOVE which is based on an open source micro-traffic simulator SUMO. MOVE allows user to quickly generate realistic mobility models for vehicular network simulations. In addition, MOVE provides an interface to automatically generate simulation scripts for ns-2 and qualnet. Finally, we show that the simulation results using MOVE is significantly different from that using the commonly used random waypoint model. Moreover, we have offered traffic light, turning, and overtaking cases studied in VANET simulation. Each case have different effect in simulation, thus selecting a suit level of detail for a simulation is key point. MOVE is publicly available and can be downloaded via the following URL - http://lens1.csie.ncku.edu.tw/MOVE/. reFerences B. A. mobility scenario generation and analysis tool, http://web.informatik.uni-bonn.de/iv/mitarbeiter/ dewaal/bonnmotion/. Bai, F., Sadagopan, N., & Helmy, A. (2003) “Important: a framework to systematically analyze the impact of mobility on performance of routing protocols for ad hoc networks,” in The 22th IEEE Annual Joint Conference on Computer Communications and Networking INFOCOM 2003 (pp. 825–835). MOVE Bai, F., Sadagopan, N., & Helmy, A. (2003, November) “The important framework for analyzing the impact of mobility on performance of routing for ad hoc networks,” AdHoc Networks Journal Elsevier Science, vol. 1, no. 4 (pp. 383–403). Bettstetter, C. (2001) “Smooth is better than sharp: a random mobility model for simulation of wireless networks,” in Proc. MSWiM01, ACM. Camp, T., Boleng, J., & Davies, V. (2002). “A survey of mobility models for ad hoc network research,” in Wireless Communications and Mobile Computing (WCMC): Special issue on Mobile Ad Hoc Networking: Research . Trends and Applications, 2(5), 483–502. Choffnes, D. R., & Bustamante, F. E. (2005) “An integrated mobility and traffic model for vehicular wireless networks,” in Proc. of the 2nd ACM International Workshop on Vehicular Ad Hoc Networks (VANET), Cologne, Germany. [Online]. Available: http://www.aqualab.cs.northwestern.edu/publications/DChoffnes05vanet.html CORSIM. (2009). http://www.fhwa-tsis.com/. CORSIM. http://mctrans.ce.ufl.edu/featured/tsis/ version5/corsim.htm. Das, S. R., Perkins, C. E., & Royer, E. M. (2000) “Performance comparison of two on-demand routing protocols for ad hoc networks,” in Proc. IEEE Infocom. David, B., & David, A. (1996) “Dynamic source routing in ad hoc wireless networks,” in Mobile Computing, vol. 35, Kluwer Academic (pp. 153–181). Floyd, S., & Paxson, V. (2001) “Difficulties in simulating the Internet,” ACM/IEEE Transactions on Networking, vol. 9, no. 4 (pp. 392–403). [Online]. Available: http://www.icir.org/vern/papers.html Haas, Z. J., & Barr, R. (2005) “Density-independent, scalable search in ad hoc networks,” in Proceedings of IEEE International Symposium on Personal Indoor and Mobile Radio Communications. [Online]. Available: http://jist.ece.cornell.edu/docs/050911density.pdf Heidemann, J., Bulusu, N., Elson, J., Intanagonwiwat, C., Lan, K. C., Xu, Y., et al. (2001) “Effects of detail in wireless network simulation,” in Proc. of Communication Networks and Distributed Systems Modeling and Simulation Conference, Jan. 2001. [Online]. Available: http://www.isi.edu/johnh/ PAPERS/Heidemann00d.html Holland, G., & Vaidya, N. H. (1999) “Proc. mobicom, acm intern. conf. on mobile computing and networking,” in Proc. IEEE Infocom. Hong, X., Gerla, M., Pei, G., & Chiang, C.-C. (1999) “A group mobility model for ad hoc wireless networks,” in Proc. ACM MSWiM. Hong, X., Kwon, T., Gerla, M., Gu, D., & Pei, G. (2001, January) “A mobility framework for ad hoc wireless networks,” in Proc. of ACM Second International Conference on Mobile Data Management (MDM 2001). Jardosh, A. P., Belding-Royer, E. M., Almeroth, K. C., & Suri, S. (2005) “Real-world environment models for mobile network evaluation,” in In the IEEE Journal on Special Areas in Communications - Special Issue on Wireless Ad hoc Networks. Johansson, P., Larsson, T., Hedman, N., Mielczarek, B., & Degermark, M. (1999) “Routing protocols for mobile ad-hoc networks - a comparative performance analysis,” in IEEE MOBICOM, p. 195V206. K. tutorial, http://www.keyhole.com/ kml/kml tut.html. Mahajan, A., Potnis, N., Gopalan, K., & Wang, A. A. (2006) “Urban mobility models for vanets,” in in Proc. of the IEEE Workshop on Next Generation Wireless Networks (WoNGeN). 367 MOVE Mangharam, R., Weller, D. S., Stancil, D. D., Rajkumar, R., & Parikh, J. S. (2005) “GrooveSim: A topography-accurate simulator for geographic routing in vehicular network,” in Proc. of the 2nd ACM International Workshop on Vehicular Ad Hoc Networks(VANET),Cologne,Germany.[Online]. Available: http://www.andrew.cmu.edu/user/rahulm/Research/Pubs/ vanet05 mangharam.pdf Marfia, G., Pau, G., Giordano, E., Sena, E. D., & Gerla, M. (2007) “Evaluating vehicle network strategies for downtown portland: Opportunistic infrastructure and importance of realistic mobility models,” in proc. of MoBiOpp 2007, co-located with ACM/USENIX International Conference on Mobile Systems, Applications, and Services (MobiSys 2007). P. simulation VISSIM. (2009). http://www.english. ptv.de/. Perkins, C., Belding-Royer, E., & Das, S. (2003) “Ad hoc on demand distance vector (aodv) routing,” rFC 3561. Perkins, C. E., & Royer, E. M. (1999) “Ad-hoc on-demand distance vector routing,” in Proc. IEEE Workshop on Mobile Computing Systems and Applications. Potnis, N., & Mahajan, A. (2006) “Mobility models for vehicular ad hoc network simulations,” in proc. of the 44th annual Southeast regional conference, ACM (pp. 746–747). Royer, E. M., Melliar-Smithy, P. M., & Mosery, L. E. (2001) “An analysis of the optimum node density for ad hoc mobile networks,” in Proc. of the IEEE International Conference on Communications, Jun. 2001. S. S. of Urban Mobility. (2009). http://sumo. sourceforge.net/. Saha, A. K., & Johnson, D. B. (2004) “Modeling mobility for vehicular ad hoc networks,” in Proc. of the 2nd ACM International Workshop on Vehicular Ad Hoc Networks (VANET). [Online]. Available: http://www.cs.rice.edu/amsaha/ Resume/ vanet2004/vanet2004.pdf 368 SimulationP. M. T. (2009). http://www.paramicsonline.com/. SimulatorO. (2009). http://www.opnet.com/. SimulatorQ. N. (2009). http://www.scalablenetworks.com/. Stepanov, I., Marron, P.-J., & Rothermel, K. (2005l) “Mobility modeling of outdoor scenarios for manets,” in Proceedings of the 38th annual Symposium on Simulation ANSS ’05, Apr. 2005 (pp. 312–322). Sven Jaap, M. B., & Wolf, L. (2005) “Evaluation of routing protocols for vehicular ad hoc networks in city traffic scenarios,” in in Proc. of the 5th International Conference on Intelligent Transportation Systems Telecommunications (ITST). T. N. S. ns 2.(2009). http://www.isi.edu/nsnam/ ns/index.html. TIGER (Topologically Integrated GEographic Encoding and Referencing). (2009). http://www. census.gov/geo/www/tiger/. TRANSIMS. http://transims.tsasa.lanl.gov/. Treiber, M., Hennecke, A., & Helbing, D. (2000) “Congested traffic states in empirical observations and microscopic simulations,” IPhys. Rev., vol. 62, no. 2. Zimmermann, H.-M., Gruber, I., & Roman, C. (2005) “A voronoi-based mobility model for urban environments,” in in Proc. Of the European Wireless 2005 (EW05) . Zonoozi, M. M., & Dassanayake, P. (1997). User mobility modeling and characterization of mobility patterns . IEEE Journal on Selected Areas in Communications, 15(7), 1239–1252. doi:10.1109/49.622908 369 Chapter 22 Security Attacks of Vehicular Networks Jen-Chun Chang National Taipei University, Taiwan, R.O.C. Chun-I Fan National Sun Yat-sen University, Taiwan, R.O.C. Ruei-Hau Hsu National Sun Yat-sen University, Taiwan, R.O.C. absTracT The application of vehicular ad hoc network (VANET) improves driving safety and traffic management. Due to the above applications, security attacks on VANET can be serious threats all the time. VANET is a special form of mobile ad hoc network (MANET). Hence any attacks exist on MANET also can be arisen on VANET. Moreover, some special attacks can be raised on VANET, which do not exist on MANET. Nevertheless, some characteristics of VANET can be positive effects and some can be negative effects on security issues. Before designing the security mechanism to defend attacks, the authors should take the positive effects and avoid the negative effects on the security of VANET. Furthermore, the authors class all possible attacks of VANET from every network layer. They also introduce the reason of forming every attack and the possible effect on VANET in detail. Therefore this chapter helps understanding the latent threats and the useful resources of security issues on VANET. inTroducTion In recent years, people have fixed their eyes upon traffic-safety topic, because current traffic accident statistics are notoriously horrific. According to the European Red Cross Road Safety Campaign report, approximately 43,000 people die every year on the roads of the European Union (EU), with around 1.8 DOI: 10.4018/978-1-60566-840-6.ch022 million people injured, and the costs associated with traffic accidents estimated 160 billion euros. The annual costs associated with crashes (like hospital bills and damaged property) total nearly 3 percent of the world’s gross domestic product (GDP) in 2000, or roughly US $1 trillion. In order to reduce the traffic accidents, governments and researchers around the world try to find out effective solutions, and Vehicular Communication (VC) system, or we can also call it Vehicular ac hoc network (VANET), Copyright © 2010, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global is prohibited. Security Attacks of Vehicular Networks seems to be one of the answers and will be ubiquitous everywhere in the not-too-distant future. Vehicular ac hoc network (VANET) is likely to become the most universal and relevant form of ad hoc networks due to the urgent need of driving safety. Another reason about the fast development of VANET is the impact to the market. There are more than 50 applications have been submitted by major car manufactures like BMW, Daimler-Chrysler, Ford, and GM which are based on Dedicated Short Range Communication (DSRC) technology. DSRC is a short range wireless protocol specifically for automotive use. It offers communication between vehicles and Road Side Units (RSUs). This technology for VANET applications is working in the 5.9 GHz band (U.S) or 5.8 GHz band (Japan, Europe). VC system includes two types of communications: Inter-vehicle communication (IVC) (or someone call it Vehicle-to-vehicle (V2V) communication) and Roadside-to-vehicle communication (RVC) (or someone call it Vehicle-to-infrastructure (V2I) communication). All two types are based on wireless multi-hop communication. inter-Vehicle communication • • • Properties that have a negative effect: • • • IVC systems have some properties that support security and others that are negative effect. Properties that have positive effects. • • 370 No energy constraints: unlike the sensor node in ad hoc network and/or sensor network and some mobile device, such as cellular phone and personal digital assistant (PDA), cars usually provide enough energy to operate communication system and related computation of security. Known position and time: This information is required for most safety application on VANET. This information can also be used to security application. Limited physical access: The operator of IVC usually limited to the owner of car or authorized personnel. Periodic maintenance: Cars always need to be maintained in a period of time. Therefore, the IVC also can check and update regularly. Secure computing platform: Automotive environment it seems inevitable that some kind of secure computing platform must be available in the future. • • High mobility: High degree of mobility is one of properties of vehicles. It means that the average speed of the node of VANET will be very high and the average connection time will be very short. Therefore, when we design the security mechanism, the communication time and computation time should be considered. Large number of nodes: IVC network can be a huge ad hoc network. Scalable solutions for adequate and sufficient performance should be considered. No centralized infrastructure: When we deal with a distributed ad hoc network, the centralized infrastructure is only available at specific situations. The design of some security building block should be adapted to such kind of infrastructure, such as trust management and key distribution and requires new concepts. Privacy concerns: Privacy is a serious problem in IVC system because cars are highly personal devices and the owners will keep it for a long duration. The system design should reflect the need for flexible identifiers. No user interaction: The scenario of IVC system is that no user interaction possible since it could distract drivers and reduce the popularity and usability of IVC system. Security Attacks of Vehicular Networks One of famous IVC system model is proposed by the network on wheels (NoW) project. The NoW system is a generic model for IVC architecture. This model consists of four major aspects. First, the radio channel and the protocol are used on NoW system. Second, the hardware and the software running on NoW. The three major kinds of platform are within the NoW system, one is onboard units (OBU), that are installed in the vehicles, road side units (RSUs) as parts of some road infrastructure, and the HomePC or commercial Access Points providing content or Internet access. Third, the input of sensor to the different processing units, that can be all kinds of physical sensors, such as temperature, oil on the road and so on. Finally, the security infrastructure behind the NoW system consists of some elements, like the vehicle manufacturers, certification authorities, traffic authorities, and certified staff. The following figure (Figure 1) is a generic system model of NoW. Roadside-to-vehicle communication (RVC) is used to complement IVC. If VANET just use IVC as communication only, all connections must be established by multi-hop communication, it is a huge works and consume large amount of resources. That is the reason we must use IVC and RVC interchange. We can also connect to Internet through RVC and get other multimedia services, like on-line TV, on-line game …etc. That make everything come true and vehicles are not just vehicles, they become to be entertainment centers. We will introduce VANET characteristics, architecture, applications and its security attack separately. VANET is a special or even exceptional case of mobile ad hoc network (MANET). MANET is one kind of wireless ad-hoc network, it is not only include the same characteristics as wireless ad-hoc network, such as an open peer-topeer network architecture, lack a central instance, each node is willing to forward data for other nodes, self-configuration and self-maintenance capabilities. Fully distributed nodes may move arbitrarily and change their topology frequently without prior notice. Figure 1. The architecture of NoW system 371 Security Attacks of Vehicular Networks The archiTecTure oF Vehicular neTwork VANET architecture is composed of two kinds of communicating nodes, including vehicles (nodes) and Road Side Units (RSUs). Vehicles can also be split into two parts: private (belonging to individuals or private companies) and public (public transportation vehicles, e.g., buses, and public services) The network architecture can roughly split into three domains (shown in Figure 2): the in-vehicle domain, the ad hoc domain and the infrastructure domain. The in-vehicle domain is a sub-network (of the VANET) with an On Board Unit (OBU) and several Application Units (AUs), such as Personal Digital Assistant (PDA), mobile phone… etc. The ad hoc domain is composed of multiple OBUs and RSUs along the road. OBUs and RSUs are equipped with IEEE 802.11-like wireless technology. They can either communicate each other directly or through multi-hop communication. Figure 2. VANET architecture 372 The infrastructure domain provides connectivity to Internet nodes. The vehicles can get applications like multimedia services through RSUs to Internet. If the vehicle we want to communicate is out of range that multi-hop can reach, we can use OBU to connect to RSU of ad hoc domain we located, and the RSU will communicate to RSU where the vehicle we want to connect located, and then we can access to the vehicle as two RSUs as intermediate nodes. The applicaTions oF Vehicular neTwork VANET could provide several safety applications (Lin et al., 2007) to protect driver and passenger and ubiquitous multimedia services. These safety applications include Electronic Emergency Barke Light (EEBL), Road Hazard Condition Notification (RHCN), Road Feature Notification (RFN), Slow Vehicle Alert (SVA), and Post Crash Notification (PCN). These safety applications could Security Attacks of Vehicular Networks Figure 3. High-level system architecture view help driver the dangerous situation or the condition of traffic to void possibly dangers or threats. EEBL notifies drivers when a vehicle is closed to them and rapidly decelerates. It could avoid the chance of accident of rear-end collisions. RHNC broadcast the information of road hazard, such as snow accumulation or trash. RFN alerts drivers the landform or sections with special limitation they will pass. If drivers approach a steep hill, school, they may need to slow their speed with a must lower ones to avoid accidents. SVA and PCN alert drivers a slow vehicle or a possible crash in the lanes ahead. The information of such safety applications are come from all nodes of VANET. Due to these safety applications, drivers have more time to reach to avoid some dangers or accidents that may happen with huge damages. The aTTacks oF Vehicular neTwork VANET is a special or even exceptional case of MANET, it includes all security attacks that MANET faced and owned. Watch out – there still exists some VANET-specific attacks. There is one thing we MUST keep in mind that security attack is cross-layer issue rather single layer issue. We can totally understand the truth from Table 1 below. The security attacks for VANETs span the entire network protocol and not just application layer VANET security attacks includes 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. Viruses, worms and malicious code Bogus information attack Message modification attack ID disclosure of other vehicles Movement tracking Denial of Services (DoS) Impersonation attack Cheating with positioning information RSU replication attack: Unauthorized preemption attack WEP vulnerability Routing attack Data packet forwarding attack 373 Security Attacks of Vehicular Networks Table 1. The security attacks for VANETs span the entire network protocol stack Layer Application Layer Security Attacks Viruses, Worms, Malicious Codes, Bogus information attack, message modification attack, Movement tracking, DoS…etc. Network Layer Ad hoc routing attack, data packet forwarding attack, ID disclosure attack, Movement tracking, DoS…etc. Data Link Layer WEP vulnerability, ID disclosure attack, Movement tracking, DoS…etc Physical Layer DoS attacks We will explain each attack behavior carefully below. Viruses, worms and malicious code: Viruses, worms and malicious code are tools used to invade OBUs or RSUs. After intrusion is successful, malicious user can get personal information of OBU owner, and convey some fake information to disturb the traffic. Therefore, the manager of VANET system can have some strategy to avoid the damages caused from virus, worms and malicious code, such as the choice of operating system, the periodic update and checking, security policy to access operating system and so on. It is also helpful to prevent such kind of threat to break VANET system by policy or standard procedure. Bogus information attack: In MANET environment, all messages are broadcast even though messages which have definitely destination for private use. Data is transmitted via air interface through shared wireless medium. It is obviously Figure 4. The in-vehicle domain 374 a loophole, and malicious user may produce fake or untrue messages to achieve certain goals. In VANET circumstance, bogus information may even cause more critical consequences, like traffic jam or traffic accidents. For example: one may send a fake traffic jam message to the other drivers and affect their behavior in order to divert traffic from a given road and get a better traffic condition (See Figure 4). Message modification attack: Some people might consider that message modification attack is similar to bogus information attack, but they are totally different. Bogus information attackers produce messages which are purely fictitious, but message modification attackers just alter rather create messages during or after transmission. The message modification attacker may wish to change the source or content of the messages. ID disclosure of other vehicles: In VANET architecture, data is transmitted via air interface. Security Attacks of Vehicular Networks And multi-hop transmission strategy makes more and more personal information is available in the infrastructure. According to the above two reasons, the probability of personal information disclosure go up quite substantially. The attackers can get users’ personal information through eavesdropping or accessing the infrastructure illegally. We call this attack technique as ID disclosure of other vehicles, also known as the Big Brother scenario. We must keep one thing firmly in mind that ID disclosure attack is cross-layer security issue. In order to authenticate client’s identity, every application service will execute an authentication protocol to client. If user lose his identity, make his identity public or eavesdropped by malicious users, attackers can masquerade by using this identity, and make victims to pay high payment bill or let them as scapegoat. The attack behavior belongs to application layer of network protocol stack. As we mentioned above, vehicular communication is based on IEEE 802.11-like technology. The applications of VANET can either use wire- less multi-hop communication or the Internet via the TCP/IP protocol stack. Every vehicle or RSU will be allocated a unique and static IP and MAC addresses by authority. The addresses represent node’s identity, and researchers try to hide the information preventing attackers capture it. If node’s IP and MAC addresses have been disclosed, user’s privacy no longer exists and user’s movement will be tracked by attackers. In Figure 5, we show header of routing packets, it includes immutable and mutable field. Immutable are those fields that unchanged from sender to destination even through intermediate hops forward the packet to the destination. Mutable fields are allocated to be altered by intermediate nodes. We can see operation on these fields in Figure 6. We can induce that immutable fields are used to store IP addresses, and mutable fields are used to store MAC addresses. The whole routing packet is not protected by encryption. Although malicious user cannot disclose identity of user from application layer message, he can still catch information Figure 5. Example of bogus information attack, attacker disseminate false information to affect the decisions of other vehicles and thus clear the way of attacker 375 Security Attacks of Vehicular Networks Figure 6. Example packet type with mutable and immutable fields (mutable fields are red, and immutable fields are blue) from routing packet. We call this attack as IP and MAC disclosure attack. movement Tracking Movement tracking goal can be done by two ways. 1. 2. Malicious user gets victim’s identity through ID disclosure attack first, and then he can proceed to track victim’s movement. Malicious user is in the area where victim’s radio range covered. As victim move, attacker also moves with the victim. If attacker never moves out of the radio range of victim, he can track victim’s movement forever. This way of movement tracking is like ‘physical’ way. Denial of service (DoS): Denial of service attack or distributed denial of service attack (DDoS attack) is an attempt to make a computer resource is unavailable to its clients. We can simply explain it as an adversary broadcasts irrelevant bulk messages to the VANET where he located in, taking up the cannel and consume the computational resource of the other nodes. There is one thing 376 we must keep in mind that DoS is a cross-layer issue, from message replay attack, layer 2 packet flooding, Radio-Frequency (RF) interference to jamming. A message replay attack means that a valid data transmission is repeated or delayed maliciously. An adversary intercepts the data from the originator and retransmits it. It may cause DoS and paralysis network or the attacker can impersonate other user successfully. Every nodes of VANET want to determine every packet’s source and destination must to dispatch packet and check MAC address or IP address. The dispatch action must take computation resource. If a node receives huge amount of packets, then he will be down due to resources are exhausted. RF (or Electromagnetic interference, EMI) is a disturbance that affects an electrical circuit due to electromagnetic radiation emitted from an external source. The disturbance may interrupt, obstruct, or otherwise degrade or limit the effective performance of the circuit. Impersonation attack: Impersonation attack is also known as Masquerade. As implied by the name, adversary actively pretends to be another vehicle or even RSU by using false identities to Security Attacks of Vehicular Networks fool the others. The attacker may try to achieve malicious or rational objectives. A malicious attacker seeks no personal benefits from the attacks and aims to harm the members or mal-function the network for no reason. That is why a malicious attacker is unpredictable, because his attack is disregarding costs and profits. All he wants is just contend to appetite for destruction. On the contrary, a rational attacker seeks personal profit and hence is more predictable in terms of the attack means and the attack target. Cheating with positioning information: Cheating with positioning information is a special case of message modification attack. The attackers just alter the positioning information field of the message like perceived position, speed, direction, and so on. Cheating with positioning information may lead to routing attack, because position-based routing protocol is the most popular routing algorithm in VANET environment. If routing table is not correct, data packet cannot be forwarded correctly, and messages may lost (e.g. sinkhole) or cause routing loops. RSU replication attack: Surveying the deployments of VANET, we can find out that there exist a large number of RSUs to complement shortcomings of multi-hop communications. If protection of RSUs is not sufficient, they might be compromised by malicious users. Afterwards, attackers can use these captured RSUs to launch any malicious attack, such as bogus information attack and unauthorized preemption attack (as we described below), etc. Maybe we can say RSU replication attack is basis of unauthorized preemption attack. Unauthorized preemption attack: The adversary may take control a RSU, especially a traffic light, and then make it to provide special traffic priority for specific vehicles. It is similar to RSU replication attack After introducing upper layer and cross-layer attacks, we will discuss lower layer (such as network and link layer) attacks below: WEP: Wired Equivalent Privacy (WEP) is an algorithm to secure IEEE 802.11 wireless networks. Because wireless networks communication is based on broadcast through radio, it is easier to eavesdrop than wired networks. That is why we need encryptosystem to provide confidentiality as wired networks, and WEP is named as this reason. But since 2001, cryptanalysts find several serious weaknesses which will let WEP to be cracked by available software about few minutes. Now we can assure that WEP is vulnerable. Routing attacks: The family of routing attack is defined as any action of notifying “wrong” routing updates (which means not follow the specifications of the routing protocol). For example, spoofing, forging of routing signaling messages …etc. are routing attacks. These behaviors may cause some specific attacks which are related to the VANET, e.g: We will introduce some specific attack behaviors which are related to the VANET, e.g.: routing loop and malfunctioning of the network, sinkhole attack…etc. We will describe in Data packet forwarding attack. Data packet forwarding attack: The attackers cause the data packets to be delivered in a way that is intentionally inconsistent with the routing states. Another type of packet forwarding attack is the Denial-of-service (DoS) attack via network-layer packet blasting, in which the attacker injects a large amount of junk packets into the network. Because ad hoc routing protocols are various, we will focus on position-based routing which is one of most popular used to construct routing table. The position–based routing protocol policy is – a node forwards a given packet to one onehop neighbor that is closest to the destination than others. Although position based routing protocol (one kind of geographical routing protocols) is common used, it is still inability to against spoofing and forging of routing signaling messages like other ad hoc routing protocols and results in creation of routing loops and mal-function of the network. 377 Security Attacks of Vehicular Networks Figure 7. Operations on mutable and immutable fields Figure 8. Two exemplary attacks in position-based routing protocol In Figure 7: The left figure shows a regular forwarding process from the source via forwarders (vehicles) V1, V2 and V3 to the destination. In the middle figure, the attacker Att forges its location as Att’. It makes V2 selects the attacker as next hop as next hop in the forwarding chain because its position is closer to destination. The attacker receives messages (that should be sent to next hop) and drops all messages. It makes all packets lost, and we call it ‘sinkhole’ attack. In the right figure, the attacker forges its position as Att’ and V2 selects the attacker as next hop in the forwarding chain. The attacker receives all messages and forwards the received messages back to the previous node V2. It results routing loop. Messages will be sent back and forth but can’t be sent to destination. 378 conclusion In this chapter, we have shown security attacks in VANET environment. We first briefly describe VANET characteristics, architectures, applications. After constructing the common sense, we elaborate all possible security attacks in VANET. Because VANET is a special case of MANET, it includes all MANET security attacks and some special attacks only happened in VANET environment. Maybe someone will ask why we need to know attacks in VANET, because VANET seems to become the most relevant network around the world and VANET includes an important component – human beings. The basis of VANET development is to protect drivers from accidents, so that VANET system must be quite reliable. How to make VANET reliable? This is quite a complicated question. But there is one thing for Security Attacks of Vehicular Networks sure that security of VANET is one of necessaries to construct a reliable VANET system. Before we design a secure system, we must find out all possible attacks first. That is what this chapter shows us. reFerences Hubaux, J. P., Capkun, S., & Luo, J. (2004). The Security and Privacy of Smart Vehicles. Security & Privacy, 2(3), 49–55. doi:10.1109/MSP.2004.26 Lin, X., Sun, X., Ho, P. H., & Shen, X. (2007). GSIS: A Secure and Privacy Preserving Protocol for Vehicular Communications. IEEE Transactions on Vehicular Technology, 56(6), 3442–3456. doi:10.1109/TVT.2007.906878 Aijz, A., Bochow, B., Dötzer, F., Festag, A., Gerlach, M., Kroh, R., & Leinmüller, T. (2006). Attacks on Inter Vehicle Communication Systems – an Analysis. In Proc. WIT. Mauve, M., Widmer, A., & Hartenstein, H. (2001). A Survey on Position-Based Routing in Mobile Ad Hoc Networks. IEEE Network, 15(6), 30–39. doi:10.1109/65.967595 Armknecht, F., Festag, A., Westhoff, D., & Zeng, K. (2007). Cross-layer Privacy Enhancement and Non-repudiation in Vehicular Communication. In 4th Workshop on Mobile Ad-Hoc Networks (WMAN). Raya, M., & Hubaux, J. P. (2005). The Security of Vehicular Ad Hoc Networks. In Proceedings of the 3rd ACM workshop on Security of ad hoc and sensor networks, (pp.11-21). Eichler, S., Dotzer, F., Schwingenscholgl, C., Javier, F., Caro, F., & Eberspacher, J, (n.d.). Secure Routing in a Vehicular Ad Hoc Network. Golle, P., Greene, D., & Staddon, J. (2004). Detecting and Correcting Malicious Data in VANETs. In Proceeding of the 1st ACM international workshop on Vehicular ad hoc networks, Session: Security in VANET, (pp.27-37). Studer, A., Luk, M., Perrig, A, (n.d.). Efficient Mechanisms to Provide Convoy Member and Vehicle Sequence Authentication in VANETs. Yang, H., Luo, H., Ye, F., Lu, S., & Zhang, L. (2004). Security in mobile ad hoc networks: Challenges and solutions. IEEE Wireless communications, 11(1), 38-47. Harsch, C., Festag, A., & Papadimitratos, P. (2007). Secure Position-Based Routing for VANETs. In IEEE 66th Vehicular Techonology Conference, (pp. 26-30). 379 380 Compilation of References 3GPP. (2008). IP Multimedia (IM) Session Handling; IM Call Model; Stage 2. TS 23.218 v8.3.0. Aarnet - australia’s research and education network. (2009). Retrieved from http://www.aarnet.edu.au/ Aguayo, D., Bicket, J., Biswas, S., Judd, G., & Morris, R. (2004). Link-level measurements from an 802.11b. Ai, Y., Sun, Y., Huang, W., & Qiao, X. (2007). OSGi based integrated service platform for automotive telematics. In Proceedings of the IEEE International Conference on Vehicular Electronic and Safety, (pp. 1-6). Aidarous, S., & Plevyak, T. (1994). Telecommunications Network Management into the 21st Century. New York: IEEE Press. Aijz, A., Bochow, B., Dötzer, F., Festag, A., Gerlach, M., Kroh, R., & Leinmüller, T. (2006). Attacks on Inter Vehicle Communication Systems – an Analysis. In Proc. WIT. Akingbehin, K. (2005). Wireless communications for intravehicle use (Wireless harnesses). Institute for Advanced Vehicle Systems, University of Michigan-Dearborn, Dearborn, MI. Akyildiz, I. F., Ho, J., & Wang, W. (1999). Mobility Management in Next-Generation Wireless Systems . Proceedings of the IEEE, 87(8), 1347–1384. doi:10.1109/5.775420 Alexopoulos, C. (1997). State Space Partitioning Methods for Stochastic Shortest Path Problems. Networks, 1(30), 9–21. doi:10.1002/(SICI)1097-0037(199708)30:1<9::AIDNET2>3.0.CO;2-H Alliance, O. M. (OMA). (2009). Retrieved from http://www. openmobilealliance.org/ AMI-C. (2002). Software API Specifications-CORE APIs. retrieved September 20 2009, from http://www.ami-c.org Amit Kumar Saha, D. B. J. (2004). Modeling mobility for vehicular ad-hoc networks. In the 1st ACM international workshop on Vehicular ad hoc networks (pp. 91–92). New York: ACM Press. Amouris, K. (2005). Position-Based Broadcast TDMA Scheduling for Mobile Ad-Hoc Networks (MANETs) with Advantaged Nodes. IEEE Military Communications Conference, 1, 252–257. Armknecht, F., Festag, A., Westhoff, D., & Zeng, K. (2007). Cross-layer Privacy Enhancement and Non-repudiation in Vehicular Communication. In 4th Workshop on Mobile Ad-Hoc Networks (WMAN). Artimy, M. (2007). Local Density Estimation and Dynamic Transmission-Range Assignment in Vehicular Ad hoc Networks. IEEE Transactions on ITS, 18(3), 400–412. Asokan, N., Kostiainen, K., Ginzboorg, P., Ott, J., & Luo, C. (2007). Applicability of identity-based cryptography for disruption-tolerant networking. In . Proceedings of MobiOpp, 07, 52–56. doi:10.1145/1247694.1247705 Assisted, G. P. S. (n.d.). Retrieved September 11, 2009, from http://en.wikipedia.org/wiki/Assisted_GPS. ASTM International E2213-03 (2003). Standard Specification for Telecommunications and Information Exchange Between Roadside and Vehicle Systems 5 GHz Band Dedicated Short Range Communications (DSRC) Medium Access Control (MAC) and Physical Layer (PHY) Specifications. B. A. mobility scenario generation and analysis tool, http:// web.informatik.uni-bonn.de/iv/mitarbeiter/dewaal/bonnmotion/. Bachir, A., & Benslimane, A. (2003). A Multicast Protocol in Ad hoc Networks Inter-Vehicle Geocast. IEEE Semiannual Vehicular Technology Conference, 4, 2456-2460. Copyright © 2010, IGI Global, distributing in print or electronic forms without written permission of IGI Global is prohibited. Compilation of References Backes, P., Tso, K., Norris, J., Tharp, G., Slostad, J., Bonitz, R., & All, K. (2000). Internet based Operations for the Mars Polar Lander Mission. In The 2000 IEEE International Conference on Robotics & Automation, (pp. 2025-2032). Bai, F., Sadagopan, N., & Helmy, A. (2003). Important: a framework to systematically analyze the impact of mobility on performance of routing protocols for ad hoc networks. In The 22th IEEE Annual Joint Conference on Computer Communications and Networking INFOCOM’03, (pp. 825–835). Balasubramanian, A., Zhou, Y., Croft, B. W., Levine, B. N., & Venkataramani, A. (2007). Web search from a bus. In Proceedings of the second workshop on Challenged networks CHANTS CHANTS ’07, (pp. 59–66). Baldwin, C. L. (2002). Designing in-vehicle technologies for older drivers: application of sensory-cognitive interaction theory. Theoretical Issues in Ergonomics Science, 3(4). doi:10.1080/1463922021000009029 Banerjee, N., Corner, M. D., & Levine, B. N. (2007). An energy-efficient architecture for dtn throwboxes. In . Proceedings of IEEE INFOCOM, 2007, 776–784. Bertsekas, D. P., & Tsitsiklis, J. N. (1991). An Analysis of Stochastic Shortest Path Problems. Mathematics of Operations Research, 16(3), 580–595. doi:10.1287/moor.16.3.580 Bettstetter, C. (2001). Smooth is better than sharp: a random mobility model for simulation of wireless networks. In Proceedings of the 4th ACM international workshop on Modeling, analysis and simulation of wireless and mobile systems (MSWIM ’01) (pp. 19–27). New York: ACM Press. Bettstetter, C., Resta, G., & Santi, P. (2003). The node distribution of the random waypoint mobility model for wireless ad hoc networks. IEEE Transactions on Mobile Computing, 2(3), 257–269. doi:10.1109/TMC.2003.1233531 Bickel, G. S. (2006). Inter/intra-vehicle wireless communication. Retrieved from http://userfs.cec.wustl.edu/~gsb1/ index.html Bicket, J., Aguayo, D., Biswas, S., & Morris, R. (2005). Architecture and evaluation of an unplanned 802.11bmesh network. In Proceedings of the 11th annual international conference on Mobile computing and networking(MOBICOM), Cologne, Germany. Available from http://www.pdos.lcs.mit. edu/papers/roofnet: mobicom05/roofnet-mobicom05.pdf Bjic, E., & Chaxel, F. (2002). Auto-ID Mobile Information System for Vehicle Life Cycle Data Management. IEEE Systems . Man and Cybernetics, 4, 6. Blaževi’c, L., Giordano, S., & LeBoudec, J. Y. (2002). Self Organized Terminode Routing. Cluster Computing Journal, 5(2), 205–208. doi:10.1023/A:1013998030317 Blum, J., Eskandarian, A., & Hoffman, L. (2004). Challenges of inter-vehicle ad hoc networks. IEEE Transactions on Intelligent Transportation Systems, 5(4), 347–351. doi:10.1109/ TITS.2004.838218 Bojarski, A., & Fichte, W. (2003). Temperature sensors. In J. Marek, H.P. Trah, Y. Suzuki,& I. Yokomori, (Ed.), Sensors Application Vol. 4 - Sensors for Automotive Technology (pp. 343-359). Berlin: Wiley-VCH. Boneh, D., & Franklin, M. (2003). Identity based encryption from the weil pairing. In SIAM Journal of Computing, 586–615. Bonnick, A. (2001). Automotive computer controlled systems. London: Butterworth Heinemann. Bononi, L., & Felice, M. D. (2007). A Cross Layered MAC and Clustering Scheme for Efficient Broadcast in VANETs. In Proceedings of the IEEE International Conference on Mobile Ad hoc and Sensor Systems. Boudec, J.-Y. L., & Vojnovic, M. (2005). Perfect simulation and stationarity of a class of mobility models. In The 24th IEEE Annual Joint Conference on Computer Communications and Networking INFOCOM’05, (pp. 2743–2754). Boyce, D. (2002). A Memoir of the ADVANCE Project. Journal of Intelligent Transportation Systems, 7(2). Boyce, D. E., Kirson, A. M., & Schofer, J. L. (1994). ADVANCE-The Illinois Dynamic Navigation and Route Guidance Demonstration Program. Advance Technology for Road Transport, IVHS and ATT. Brakatsoulas, S., Pfoser, D., Salas, R., & Wenk, C. (2005). On Map-Matching Vehicle Tracking Data. In The 31st VLDB Conference, (pp. 853 – 864). Brewer, E., Demmer, M., Du, B., Ho, M., Kam, M., & Nedevschi, S. (2005). The case for technology in developing regions. IEEE Computer, 38, 25–38. 381 Compilation of References Briesemeister, L., & Hommel, G. (2000). Overcoming Fragmentation in Mobile Ad hoc Networks. Journal of Communications and Networks, 2(3), 182–187. Broggi, A., Bertozzi, M., Fascioli, A., Guarino, C., Bianco, L., & Piazzi, A. (1999). The Argo Autonomous Vehicle’s Vision And Control Systems. International Journal of Intelligent Control and Systems, 3, 409–441. Brown, A., Kolberg, M., Bushmitch, D., Lomako, G., & Ma, M. (2006). A SIP-based OSGi Device Communication Service for Mobile Personal Area Networks. In Consumer Communications and Networking Conference(CCNC). Buede, D. M. (2000). The engineering and design of systems: Models and methods. New York: Wiley, Inc. Burgess, J., Gallagher, B., Jensen, D., & Levine, B. N. (2006). Maxprop: Routing for vehicle-based disruptiontolerant networks. In . Proceedings of IEEE INFOCOM, 2006, 1–11. doi:10.1109/INFOCOM.2006.228 Burns, B., Brock, O., & Levine, B. N. (2005). Mv routing and capacity building in disruption tolerant networks. Proceedings of IEEE INFOCOM, 1, 398–408. Bushmitch, D. Lin, Wanrong., Bieszczad, A., Kaplan, A., Papageorgiou, V., & Pakstas, A. (2004). A SIP-based device communication service for OSGi framework. IEEE Consumer Communications and Networking Conference2006. Caliskan, M., Graupner, D., & Mauve, M. (2006). Decentralized discovery of free parking places. In Proceedings of the 3rd International ACM Workshop on Vehicular ad hoc networks, VANET’06, (pp. 30–39). New York: ACM Press. Camp, J., Robinson, J., Steger, S., & Knightly, E. (2006). Measurement driven deployment of a two-tier urban mesh access network. In Proceedings of ACM MobiSys 2006, Uppsala, Sweden. Available from http://networks.rice.edu/ papers/sys7122-camp.pdf Camp, T., Boleng, J., & Davies, V. (2002). A survey of mobility models for ad hoc network research. Wireless Communications and Mobile Computing (WCMC): Special issue on Mobile Ad Hoc Networking: Research . Trends and Applications, 2(5), 483–502. Chaintreau, A., Hui, P., Crowcroft, J., Diot, C., Gass, R., & Scott, J. (2005). Pocket switched networks: Real-world mobility and its consequences for opportunistic forward- 382 ing. Technical report, University of Cambridge Computer Laboratory, Cambridge, UK. Chakravorty, R., & Pratt, I. (2002). Performance issueswith general packet radio service. Journal of Communications and Networks (JCN), 4(2). Available from http://www.cl.cam. ac.uk/Research/SRG/netos/papers/comob-web/2002-jcn. pdf Chan, Alex Y. M., & Lu, W. P. (2003). Architecture for Wireless Access in Vehicles. IEEE Vehicular Technology Conference, 5, 3336-3340. Chang, B.-J., Huang, B.-J., & Liang, Y.-H. (2008). Wireless Sensor Network-based Adaptive Vehicle Navigation in Multihop-Relay WiMAX Networks. The 22nd International Conference on Advanced Information Networking and Applications, (pp. 56-63). Chen, A., Jain, N., Perinola, A., Pietraszek, T., Rooney, S., & Scotton, P. (2004). Scaling Real-Time Telematics Applications using Programmable Middleboxes: A Case Study in Traffic Prediction. In First IEEE Consumer Communications and Networking Conference, (pp. 388-393). Chen, J. (1998). A Study of Web-Based SNMP Network Management with a Simple Java Applet Network Monitoring Tool. Department of Computer Science and Engineering Auburn University, Alabama. Chen, J. L., Chang, Y. C., Lin, Y. S., & Du, H. W. (2009). Embedded WiMAX-based Vehicular Router for Telematics Computing. In The 11th International Conference on Advanced Communication Technology, (pp. 1857-1862). Chen, Y. S., Lin, Y. W., & Lee, S. L. (2009). A Mobicast Routing Protocol for Vehicular Ad Hoc Networks. Mobile Networks and Applications (MONET).New York: ACM/ Springer. Chen, Y. S., Lin, Y. W., & Pan, C. Y. (2009). A DiagonalIntersection-Based Routing Protocol for Urban Vehicular Ad Hoc Networks. Telecommunication System, (under revision). Chen, Y., Zhao, W., Ammar, M., & Zegura, E. (2007). Hybrid routing in clustered dtns with message ferrying. In Proceedings of the 1st international MobiSys workshop on Mobile opportunistic networking MobiOpp ’07, (pp. 75–82). Chisalita, I., & Shahmehri, E. (2004). Vehicular communication: A candidate technology for traffic safety. In IEEE Compilation of References International Conference on Systems, Man and Cybernetics, (pp. 3903–3908). Choffnes, D. R., & Bustamante, F. E. (2006). An integrated mobility and traffic model for vehicular wireless networks. In The 2nd ACM international workshop on Vehicular ad hoc networks, VANET’05, (pp. 69–78). New York: ACM Press. Choi, D. H., Ro, S. H., Lee, K. S., Lee, W. N., Choi, S. Y., & Pae, Y. W. (2006). Telematics Applications Based On TOPAZ Platform. In IEEE 6th International Conference on Computer and Information Technology, (pp.262-268). Choi, J. W., Han, W. Y., Kim, C. S., & Kwon, O. C. (2005). Open Telematics Services Deployment on the Gateway and Framework Independent on Mobile Networks. In International Conference on Wireless Networks, (pp.374-379). Chuah, M., & Xi, Y. (2007). An encounter-based multicast scheme for disruption tolerant networks. Technical report published by the University of Lehigh, Lehigh, PA. Connexis website. (n.d.). Retrieved September 20, 2009, from http://www.connexis.com/ CORSIM. (2009). http://www.fhwa-tsis.com/. CORSIM. http://mctrans.ce.ufl.edu/featured/tsis/version5/ corsim.htm. Damn small linux not (dsl-n), (n.d.). Retrieved from http:// www.damnsmalllinux.org/dsl-n/ Das, S. R., Perkins, C. E., & Royer, E. M. (2000) “Performance comparison of two on-demand routing protocols for ad hoc networks,” in Proc. IEEE Infocom. David, B., & David, A. (1996). Dynamic source routing in ad hoc wireless networks. In Mobile Computing, (vol. 35, pp. 153–181). Kluwer Academic. Deflorio, F. P. (2003). Evaluation of a reactive dynamic route guidance strategy. Journal of the Transportation Research Board Transportation, 11(5), 375–388. Demers, A. A., Greene, D., Hauser, C., Irish, W., Larson, J., Shenker, S., et al. (1987). Epidemic algorithms for replicated database maintenance. In Proceedings of the Sixth Symposium on Principles of Distributed Computing, (pp. 1–12). Demmer, M., & Ott, J. (2006). Delay tolerant networking tcp convergence layer protocol. Internet-draft. Devarapalli, V., Wakikawa, R., Petrescu, A., Thubert, P. (2005). Network Mobility (NEMO) Basic Support Protocol RFC 3963. Differential, G. P. S. (n.d.). Retrieved September 11, 2009, from http://en.wikipedia.org/wiki/Differential_GPS. Dijkstra, E. W. (1959). A note on two problems in connection with graphs. Nuerische Mathematik, 1, 269–271. doi:10.1007/BF01386390 Dikajakos, M. D., Florides, A., Nadeem, T., & Iftode, L. (2008). Location-Aware Services over Vehicular Ad-Hoc Networks using Car-to-Car Communications. IEEE Journal on Selected Areas in Communications, 25(8), 1590–1602. doi:10.1109/JSAC.2007.071008 DLNA. (2009). Retrieved from http://www.dlna.org/ home Draft 7.0, IEEE Standard 1609.1 (2006). IEEE Trial-Use Standard for Wireless Access in Vehicular Environments (WAVE) – Resource Manager. Drane, C., & Rizos, C. (1998). Positioning system in intelligent transport systems. Boston: Artech House. Dreyfus, S. E. (1969). An Appraisal of Some Shortest Path Algorithms. Operations Research, 17, 395–412. doi:10.1287/ opre.17.3.395 Duri, S., Gruteser, M., Liu, X., Moskowitz, P., Perez, R., Singh, M., & Tang, J. M. (2002). Framework for security and privacy in automotive telematics. In Mobile Computing and Networking, (pp. 25-32). Eichler, S. (2007). Performance evaluation of the IEEE 802.11p WAVE communication standard. In Proceedings of IEEE Vehicular Technology Conference, (pp. 2199-2203). Eichler, S., Dotzer, F., Schwingenscholgl, C., Javier, F., Caro, F., & Eberspacher, J, (n.d.). Secure Routing in a Vehicular Ad Hoc Network. EI-Rabbany. A. (2002). Introduction to GPS – The Global Positioning System. Boston: Artech House. Emst, T. (2006). Network Mobility Support Goals and Requirements. Encyclopedia.com. (2009). Telematics is not a question of if, but when. Retrieved from http://www.encyclopedia.com/ doc/1G1-94875067.html 383 Compilation of References Ernst, T., Uehara, K., & Mitsuya, K. (2003). Network Mobility from the InternetCAR Perspective. In 17th International Conference on Advanced Information Networking and Applications, (pp. 19–26). ETSI TISPAN. (2008). Retrieved from http://www.etsi. org/tispan/ Fall, K. (2003). A delay-tolerant network architecture for challenged internets. In Proceedings of ACM SIGCOMM, (pp. 27–34). Farrell, J. A. (2008). Aided Navigation: GPS With High Rate Sensors (pp. 22-80). New York: Glencoe/McGrawHill School Pub Co. Farrell, S., & Cahill, V. (2005). Ltp-t: A generic delay tolerant transport protocol. Technical report, University of Dublin, Dublin. Farrell, S., Ramadas, M., & Burleigh, S. (2007). Licklider transmission protocol - extensions. Internet-draft. Fasbender, A., Gerdes, M., Hjelm, J., Kvarnström, B., Petersson, J., & Skog, R. (2008). Virtually at home: High-performance access to personal media. Ericsson Review, 2. munications and navigation system for multiple platforms. In Proceedings of the IEEE/MTS OCEANS Conference, Washington DC, (Vol. 2, pp. 1086–1092). Friesz, T. L., Luque, J., Tobin, R. L., & Byung-Wook, W. (1989). Dynamic network traffic assignment considered as a continuous time optimal control problem. Operations Research, 37(6), 893–901. doi:10.1287/opre.37.6.893 Frost & Sullivan, (2005). Strategic Analysis of the Taiwan Telematics and Infotainment Market. Fu, L. (2001). Adaptive Routing Algorithm for In-Vehicle Route Guidance Systems with Real-Time Information. Transportation Research Part B: Methodological, 35(8), 749–765. doi:10.1016/S0191-2615(00)00019-9 Fukuhara, T., Warabino, T., Ohseki, T., Saito, K., Sugiyama, K., Nishida, T., & Eguchi, K. (2005). Broadcast Methods for Inter-Vehicle Communications System. In Proceedings of the IEEE Wireless Communications and Networking Conference. Feit, S. (1993). TCP/IP: Architecture, protocols and implementation. New York: McGraw Hill, Inc. Ganguly, S., Navda, V., Kim, K., Kashyap, K., Niculescu, D., Izmailov, R., et al. (2006). Performance optimizations for deploying voip services in mesh networks. Retrieved from http://www.wings.cs.sunysb.edu/%7Eanand/papers/ jsac06.pdf Feit, S. (1995). SNMP: A Guide To Network Management. New York: McGraw Hill, Inc. Geer, D. (2005). Users Make a Beeline for ZigBee Sensor Technology. Computer, 16–19. Feit, S. (1995). SNMP: A guide to network management. New York: McGraw Hill, Inc. Gerbers, J. (2003). High-pressure sensors. In J. Marek, H.P. Trah, Y. Suzuki,& I. Yokomori, (Ed.), Sensors Application Vol. 4 - Sensors for Automotive Technology (pp. 333-342). Berlin: Wiley-VCH. Fleming, W. J. (2008). New automotive sensors−A review. IEEE Sensors Journal, 8(11), 1900–1921. doi:10.1109/ JSEN.2008.2006452 FlexRay Communications System Protocol Specification v2.1 Revision A. (2005). Gerlach, M., & Fokus, F. (2007). Trust for Vehicular Applications. In 8th International Symposium on Autonomous Decentralized Systems, (pp. 295-304). FlexRay Consortium. (2009). Retrieved from http://www. flexray.com Gerten, G. (2005). Protecting the Global Positioning System. IEEE Aerospace and Electronic Systems Magazine, 20(11), 3–8. doi:10.1109/MAES.2005.1576067 Floyd, S., & Paxson, V. (2001) “Difficulties in simulating the Internet,” ACM/IEEE Transactions on Networking, vol. 9, no. 4 (pp. 392–403). [Online]. Available: http://www.icir. org/vern/papers.html Global Positioning System. (n.d.). Retrieved September 11, 2009, from http://en.wikipedia.org/wiki/Global_Positioning_System. Freitag, L., Grund, M., Singh, S., Partan, J., Koski, P., & Ball, K. (2005). The whoi micro-modem: An acoustic com- 384 Goel, A., Ramakrishnan, K. G., Kataria, D., & Logothetis, D. (2001). Efficient Computation of Delay-Sensitive Routes from One Source to All Destinations. IEEE Conference on Computer Communications (INFOCOM), (pp. 854–858). Compilation of References Golle, P., Greene, D., & Staddon, J. (2004). Detecting and Correcting Malicious Data in VANETs. In Proceeding of the 1st ACM international workshop on Vehicular ad hoc networks, Session: Security in VANET, (pp.27-37). Gong, Y., Xiong, Y., Zhang, Q., Zhang, Z., Wang, W., & Xu, Z. (2006). Anycast routing in delay tolerant networks. In IEEE GLOBECOM (pp. 1–5). Gorgorin, C., Gradinescu, V., Diaconescu, R., Cristea, V., & Ifode, L. (2006). An integrated vehicular and network simulator for vehicular ad-hoc networks. In the 20th European Simulation and Modelling Conference. Green, P., Levison, W., Paelke, G., & Serafin, C. (1995). Preliminary Human Factors Design Guidenlines for Driver Information Systems. Technical Report No. UMTRI-93-21, Transportation Research Institute, The University of Michigan, Ann Arbor, MI. Grossglauser, M., & Tse, D. N. C. (2002). Mobility increases the capacity of ad-hoc wireless networks. In IEEE/ACM Transactions on Networking, 10(4), 477–486. Grymek, L., Singh, S., & Pattipati, K. (2007). Vehicular Dependence Adds to Telematics’ Allure. IEEE Potentials, 26(2), 12–16. doi:10.1109/MP.2007.343053 Gsmworld.com. (2009). Retrieved from http://www.gsmworld.com/technology/gprs/index.shtml Gun, N., Held, A., & Kaiser, J. (2001). Proactive services in a distributed traffic telematics application. International Workshop on Mobile Communication over Wireless LAN: Research and Applications. Guo, J. H., & Xing, G. M. (2007). Using Mobile Agentbased Middleware to Support Distributed Coordination for Vehicle Telematics. In 21st International Conference on Advanced Information Networking and Applications Workshops, (Vol. 2, pp.374-379). Gura, N., Held, A., & Kaiser, J. (2001). Proactive Services in a Distributed Traffic Telematics Application. Workshop on Mobile communication over Wireless LAN: Research and Applications. Retrieved from http://www.informatik. uni-ulm.de/rs/projekte/core/ Haas, Z. J., & Barr, R. (2005) “Density-independent, scalable search in ad hoc networks,” in Proceedings of IEEE International Symposium on Personal Indoor and Mobile Radio Communications. [Online]. Available: http://jist.ece. cornell.edu/docs/050911-density.pdf Hackbarth, K. (2003). OSGi – Service-Delivery-Platform for Car Telematics and Infotainment Systems. Advanced Microsystems for Automotive Applications 2003 (pp. 497507). Berlin: Springer. Hall, R. (1986). The Fastest Path through a Network with Random Time- Dependent Travel Time. Transportation Science, 20(3), 182–188. doi:10.1287/trsc.20.3.182 Han, W. Y., Kwon, O. C., Park, J. H., & Kang, J. H. (2005). A Gateway and Framework for Interoperable Telematics Systems Independent on Mobile Networks. ETRI Journal, 27(1), 106–109. doi:10.4218/etrij.05.0204.0038 Harsch, C., Festag, A., & Papadimitratos, P. (2007). Secure Position-Based Routing for VANETs. In IEEE 66th Vehicular Techonology Conference, (pp. 26-30). Heidemann, J., Bulusu, N., Elson, J., Intanagonwiwat, C., Lan, K. C., Xu, Y., et al. (2001) “Effects of detail in wireless network simulation,” in Proc. of Communication Networks and Distributed Systems Modeling and Simulation Conference, Jan. 2001. [Online]. Available: http://www.isi.edu/ johnh/PAPERS/Heidemann00d.html Holland, G., & Vaidya, N. H. (1999) “Proc. mobicom, acm intern. conf. on mobile computing and networking,” in Proc. IEEE Infocom. Home Gateway Initiative. (2007). Retrieved from http:// www.homegatewayinitiative.org/ Hong, X., Gerla, M., Pei, G., & Chiang, C.-C. (1999) “A group mobility model for ad hoc wireless networks,” in Proc. ACM MSWiM. Hong, X., Kwon, T., Gerla, M., Gu, D., & Pei, G. (2001, January) “A mobility framework for ad hoc wireless networks,” in Proc. of ACM Second International Conference on Mobile Data Management (MDM 2001). Hsu, R. C., & Chen, L. R. (2005). Integrated Embedded System Architecture for In-Vehicle Telematics and Infotainment System. Proceedings of the IEEE International Symposium on Industrial Electronics, 4, 1409–1414. Hubaux, J. P., Capkun, S., & Luo, J. (2004). The Security and Privacy of Smart Vehicles. Security & Privacy, 2(3), 49–55. doi:10.1109/MSP.2004.26 385 Compilation of References Hughes, L., & Maghsoudlou, A. (2006). An Efficient Coverage-Based Flooding Scheme for Geocasting in Mobile Ad Hoc Networks. International Conference on Advanced Information Networking and Applications, 1, 6. Hyyrylainen, T., Karkkainen, T., & Luo, C. (2007). Opportunistic email distribution and access in challenged heterogeneous environments. In Proceedings of the second workshop on Challenged networks CHANTS (pp.97–100). Ibm.com. (n.d.). Focus on the road ahead: IBM puts practical telematics within reach. Retrieved from http:// www-306.ibm.com/software/pervasive/info/Telematics_within_reach_050404.pdf IEEE P802.11p/D3.0. (2007). Draft amendment for wireless access in vehicular environments (WAVE). IEEE Standard 1609.3 (2007). IEEE Trial-Use Standard for Wireless Access in Vehicular Environments (WAVE) – Networking Services. IEEE Standard 1609.4 (2006). IEEE Trial-Use Standard for Wireless Access in Vehicular Environments (WAVE) – Multi-channel Operation. IEEE Std 1488-1999 (2000). IEEE Standard for Message Set Template for Intelligent Transportation Systems. IEEE Std 1489-1999 (1999). IEEE Standard for Data Dictionaries for Intelligent Transportation Systems. ment 2: Physical and medium access control layers for combined fixed and mobile operation in licensed bands and corrigendum 1. IEEEP802.11p (2009, May). IEEE Draft Standard for Information Technology – Telecommunications and information exchange between systems – Local and metropolitan area networks – Specific requirements – Part 11: Wireless LAN Medium Access Control (MAC) and Physical Layer (PHY) specifications Amendment 7: Wireless Access in Vehicular Environments. IEEEStandard1609.2 (2006). IEEE Trial-Use Standard for Wireless Access in Vehicular Environments (WAVE) – Security Services for Applications and Management Messages. IETF Network Working Group. (2008). RFC3344- IP Mobility Support for IPv4. Retrieved from http://www.faqs. org/ rfcs/rfc3344.html InCode Telecom Group Inc. (2009). Telematics: How economic and technological forces will shape the industry in the U.S. Retrieved from http://www.incodewireless.com/ pdfMailer/files/Telematics_Position_Paper_v11.pdf Institute of Transportation Engineers Management and Operations of Intelligent Transportation Systems. (2003). ITS standards overview. IEEE Std 828-1998. (1998). IEEE standard for software configuration management plans. Internet Librarian, D. I. S. A. (1997). US-DOD Internet related standardized profiles. Retrieved from http://wwwlibrary .itsi.disa.mil/ IEEE1609.1TM (2006). IEEE Trial-Use Standard for Wireless Access in Vehicular Environments (WAVE) – Resource Manager. Internet, C. A. R. Project (2003). InternetCAR webpage at WIDE. Retrieved September 20, 2009 from http://www.sfc. wide.ad.jp/InternetCAR IEEE1609.2TM (2006). IEEE Trial-Use Standard for Wireless Access in Vehicular Environments (WAVE) – Security Services for Applications and Management Message. Interplanetary internet project (n.d.). Interplanetary internet project. Retrieved from http://www.ipnsig.org IEEE1609.3 (2006). IEEE Trial-Use Standard for Wireless Access in Vehicular Environments (WAVE) – Networking Services. TM IEEE1609.4TM (2006). IEEE Trial-Use Standard for Wireless Access in Vehicular Environments (WAVE) – Multichannel Operation. IEEE802.16-2005. (2006). Part 16: Air interface for fixed and mobile broadband wireless access systems amend- 386 Ip, Y. K., Lau, W. C., & Yue, O. C. (2007). Forwarding and replication strategies for dtn with resource constraints. In . Proceedings of IEEE Vehicular Technology Conference, 1, 1260–1264. ITS Standards Outreach, Education and Training Program, Institute of Transportation Engineers Management and Operations of Intelligent Transportation Systems (2003). ITS Standards Overview. Compilation of References ITS Standards Outreach, Education and Training Program, Institute of Transportation Engineers. (2006). Center to center communications. ITU-T X.680-X.690 (1994). ISO/IEC 8824: Abstract Syntax Notation One (ASN.1). Jain, S., Fall, K., & Patra, R. (2004). Routing in a delay tolerant network. In Proceedings of ACM SIGCOMM, (Vol. 34, pp. 145–158). Jakoby, B., & Herrmann, F. (2003). Chemical sensors for liquid media. In J. Marek, H.P. Trah, Y. Suzuki,& I. Yokomori, (Ed.), Sensors Application Vol. 4 - Sensors for Automotive Technology (pp. 516-526). Berlin: Wiley-VCH. Jardosh, A. P., Belding-Royer, E. M., Almeroth, K. C., & Suri, S. (2005) “Real-world environment models for mobile network evaluation,” in In the IEEE Journal on Special Areas in Communications - Special Issue on Wireless Ad hoc Networks. Jardosh, A., Ramachandran, K., Almeroth, K. C., & Belding-Royer, E. M. (2005). Understanding congestion in ieee 802.11b wireless networks. In Proceeding of ACM SIGCOMM Internet Measurement Conference, Berkeley, CA. Retrieved from http://moment.cs.ucsb.edu/amitj/ jardosh-imc2005.pdf Jiang, D., & Delgrossi, L. (2008). IEEE 802.11p: Towards an international standard for wireless access in vehicular environments. In IEEE Vehicular Technology Conference, (pp. 2036-2040). Johansson, K. H., Torngren, M., & Nielsen. L. (2003). Vehicle applications of controller area network. Technical Report Department of Signals, Sensors and Systems, Royal Institute of Technology, Stockholm, Sweden, Department of Electrical Engineering, Linkoping University, Sweden. Johansson, P., Larsson, T., & Hedman, N. (1999). Scenariobased performance analysis of routing protocols for mobile ad-hoc networks. In MOBICOM 99, The Fifth Annual ACM/ IEEE International Conference on Mobile Computing and Networking, Seattle, WA (pp. 15–19). Johansson, P., Larsson, T., Hedman, N., Mielczarek, B., & Degermark, M. (1999) “Routing protocols for mobile ad-hoc networks - a comparative performance analysis,” in IEEE MOBICOM, p. 195V206. K. tutorial, http://www.keyhole. com/kml/kml tut.html. Joshi, H. P., Sichitiu, M., & Kihl, M. (2007). Distributed Robust Geocast Multicast Routing for Inter-Vehicle Communication. WEIRD Workshop on WiMax, Wireless and Mobility. Jost, D., & Nagel, K. (2003). Probabilistic traffic flow breakdown in stochastic car following models. Transportation Research Record, 1852, 152–158. doi:10.3141/1852-19 Jun, H., Ammar, M. H., & Zegura, E. W. (2005). Power management in delay tolerant networks: A framework and knowledge-based mechanisms. In IEEE Communications Society Conference on Sensor and Ad Hoc Communications and Networks (SECON), (pp. 418–429, 2005). Jun, H., Ammar, M. H., Corner, M. D., & Zegura, E. W. (2006). Hierarchical power management in disruption tolerant networks with traffic-aware optimization. In Proc. ACM SIGCOMM Workshop on Challenged Networks (CHANTS), (pp. 245–252). Jung, C. R., & Kelber, C. R. (2005). Lane following and lane departure using a linear-parabolic model. Image and Vision Computing, 23, 1192–1202. doi:10.1016/j.imavis.2005.07.018 Juo, C. S., & Pan, J. Y. (2008). Software Agent Framework for Dynamic Handoff Decision. In APCC 2008, Akihabara, Tokyo, Japan. Ergen, M. (2009). The Access Service Network in WiMAX: The Role of ASN-GW. Kaplan, E. D., & Hegarty, C. J. (2006). Understanding GPS – Principle and Applications. Boston: Artech House. Karimi, A., Hegyi, A., Schutter, B. D., Hellendoorn, J., & Middelham, F. (2004, Oct.). Integrated model predictive control of dynamic route guidance information systems and ramp metering. Paper presented at the Proceedings of the 7th International IEEE Conference on Intelligent Transportation Systems (ITSC 2004), Washington, DC. Karnadi, F. K., Mo, Z. H., & Lan, K. C. (2007). Rapid generation of realistic mobility models for vanet. In IEEE Wireless Communications and Networking Conference. Karp, B. (2001). Challenges in Geographic Routing: Sparse Networks, Obstacles, and Traffic Provisioning. In DIMACS Workshop on Pervasive Networking. Karp, B., & Kung, H. T. (2000). GPSR: Greedy Perimeter Stateless Routing for Wireless Networks. ACM International Conference on Mobile Computing and Networking, (pp. 243-254). 387 Compilation of References Kastinaki, V., Zervakis, V., & Kalaitzakis, K. (2003). A survey of video processing techniques for traffic applications. Image and Vision Computing, 21(4), 359–381. doi:10.1016/ S0262-8856(03)00004-0 Kim, J. H., Seo, S. H., Moon, T. Y., Hwang, S. H., & Jeon, J. W. (2007). A method of improving the reliability of the gateway system by using OSEK/VDX. In Proceedings of IEEE International Conference on Control, Automation and System, (pp. 2328-2833). Kim, M. J., Choi, Y. J., Moon, Y. B., Kim, S. J., & Kwon, O. C. (2006). Design and Implementation of Status based Application Manager for Telematics. IEEE 8th International Conference on Advanced Communication Technology, (pp. 20-22). Kim, T., Kim, D., Park, N., Yoo, S., & Lopez, T. S. (2007). Shortcut tree routing in ZigBee networks. In . Proceedings of ISWPC, 07, 42–47. Kofink, P. (2003). Steering-angle sensors. In J. Marek, H.P. Trah, Y. Suzuki,& I. Yokomori, (Ed.), Sensors application Vol. 4 - Sensors for automotive technology (pp. 428-449). Berlin: Wiley-VCH. Korkmaz, G., Ekici, E., & Ozguner, F. (2006). A CrossLayer Multihop Data Delivery Protocol with Fairness Guarantees for Vehicular Networks. IEEE Transactions on Vehicular Technology, 55(3), 865–875. doi:10.1109/ TVT.2006.873838 Kotsialos, A., Papageorgiou, M., Diakaki, C., Pavlis, Y., & Middelham, F. (2002). Traffic flow modeling of largescale motorway networks using the macroscopic modeling tool METANET. Journal of IEEE Transaction on Intelligent Transportation Systems, 3(4), 282–292. doi:10.1109/ TITS.2002.806804 Krotkov, E., & Blitch, J. (1999). The defense advanced research projects agency (darpa) tactical mobile robotics program. In The International Journal of Robotics Research (pp. 769–776). Kuhne, R. D., & Langbein-Euchner, K. (1995). Calculation of travel time savings by dual mode route guidance for the South corridor in the Stuttgart test field. Vehicle Navigation and Information Systems Conference. LeBrun, J., Chuah, C.-N., Ghosal, D., & Zhang, M. (2005). Knowledgebased opportunistic forwarding in vehicular 388 wireless ad hoc networks. In IEEE Vehicular Technology Conference (VTC), (pp. 2289– 2293). Lee, C., Nordstedt, D., & Helal, S. (2003). Enabling Smart Spaces with OSGi. IEEE Pervasive Computing, 2(3, JulySept), 89-94. Lee, J., Forlizzi, J., & Hudson, S. E. (2005). Studying the effectiveness of MOVE: a contextually optimized in-vehicle navigation system. In Conference on Human Factors in Computing Systems. Lee, K. K., Kim, S. H., Choi, Y. S., & Park, H. S. (2006). A mesh routing protocol using cluster label in the ZigBee network. In Proceedings of Mobile Ad-hoc and Sensor Systems (MASS), (pp. 801-806). Lee, S. W., Munson, J., Lee, D. R., Thompson, G., & Park, J. (2005). A Rule-based System for Sense-and-Respond Telematics Applications. In The First Korea/Japan Joint Workshop on Ubiquitous Computing & Networking Systems, (pp 109-114). Lee, W., Tak, Y., Byun, W. H., & Pae, Y. W. (2005). Toward an Open Telematics Infrastructure. IPSJ SIG Technical Reports, (pp. 443-448). Leguay, J., Friedman, T., & Conan, V. (2006). Evaluating mobility pattern space routing for DTNs. In Proceedings of IEEE INFOCOM. Leick, A. (2004). GPS Satellite Surveying. New York: John Wiley & Sons. Li, Y., Wang, F., Feng, H., & Li, Z. (2005). OSGi-based service gateway architecture for intelligent automobiles. In Proceedings of the IEEE International Conference on Vehicles Symposium, (pp. 861-865). Ligs, J. F., & Bowcott, S. (1995). ADVANCE: Initial Deployment. Annual Meeting of ITS America. Lin, X., Sun, X., Ho, P. H., & Shen, X. (2007). GSIS: A Secure and Privacy Preserving Protocol for Vehicular Communications. IEEE Transactions on Vehicular Technology, 56(6), 3442–3456. doi:10.1109/TVT.2007.906878 Lindgren, A., & Doria, A. (2007). Experiences from deploying a reallife dtn system. In 2007 4th IEEE Consumer Communications and Networking Conference, (pp. 217–221). Lindgren, A., Doria, A., & Schelen, O. (2003). Probabilistic routing in intermittently connected networks. ACM SIG- Compilation of References MOBILE Mobile Computing and Communications Review, 7(3), 19–20. doi:10.1145/961268.961272 Lochert, C., Mauve, M., Füßler, H., & Hartenstein, H. (2005). Geographic routing in city scenarios. ACM SIGMOBILE Mobile Computing and Communications, 9(1), 69–72. doi:10.1145/1055959.1055970 Locust world. (2009). Retrieved from http://www.locustworld.com Lundgren, H., Ramachandran, K., Belding-Royer, E., Almeroth, K., Benny, M., Hewatt, A., Touma, A., & Jardosh, (2006). Experiences from building and using the ucsb meshnet testbed. IEEE Wireless network. Available from http://user.it.uu.se/[UNKNOWN ENTITY &tildent;] henrikl/publications.html mesh network. Luo, J., & Hubaux, J.-P. (2004). A survey of inter-vehicle communication. School of computer and Communication Sciences, EPEL, Tech. Rep. IC/2004/24. Machen. L. (2009). Technical marketing engineer, “Intel drives in-vehicle solutions” handheld components division. Retrieved from http://www.intel.com/technology/magazine/ computing/it04001.pdf Mahajan, A., Potnis, N., Gopalan, K., & Wang, A.-I. A. (2006). Urban mobility models for vanets. In Proceedings of the 2nd IEEE International Workshop on Next Generation Wireless Networks. Mahfoud, M., Al-Holou, N., & Baroody, R. (2008). Next generation vehicle network: Web enable. In Proceedings of 3rd International Conference on Information and Communication Technologies: From Theory to Applications, (pp. 1-7). Mangharam, R., Weller, D. S., Stancil, D. D., Rajkumar, R., & Parikh, J. S. (2005) “GrooveSim: A topography-accurate simulator for geographic routing in vehicular network,” in Proc. of the 2nd ACM International Workshop on Vehicular Ad Hoc Networks(VANET),Cologne,Germany.[Online]. Available: http://www.andrew.cmu.edu/user/rahulm/Research/Pubs/ vanet05 mangharam.pdf Manoharan, R., & Thambidurai, S. L. P. P. (2008). Energy Efficient Robust On-Demand Multicast Routing Protocol for MANETs. [IJAHUC]. International Journal of Ad Hoc and Ubiquitous Computing, 3(2), 90–98. doi:10.1504/ IJAHUC.2008.017002 Marfia, G., Pau, G., Giordano, E., Sena, E. D., & Gerla, M. (2007) “Evaluating vehicle network strategies for downtown portland: Opportunistic infrastructure and importance of realistic mobility models,” in proc. of MoBiOpp 2007, co-located with ACM/USENIX International Conference on Mobile Systems, Applications, and Services (MobiSys 2007). Mariño, P., Poza, F., Dominguez, M. A., & Otero, S. (2009). Electronics in automotive engineering: A top-down approach for implementing industrial field bus technologies in city buses and coaches. IEEE Transactions on Industrial Electronics, 589–600. doi:10.1109/TIE.2008.2002723 Marples, D., & Kriens, P. (2001). The Open Services Gateway Initiative: An Introduction Overview. IEEE Communications Magazine, 39(12), 110–114. doi:10.1109/35.968820 Mauve, M., Widmer, A., & Hartenstein, H. (2001). A Survey on Position-Based Routing in Mobile Ad Hoc Networks. IEEE Network, 15(6), 30–39. doi:10.1109/65.967595 Messmer, & Papageorgiou, M. (1990). METANET: A macroscopic simulation program for motorway networks. Journal of Traffic Engineering & Control, 31(8-9), 466-470. Michael, A. (2004). Guide to the IEEE 1512 family of standards. Washington, DC: Institute of Electrical & Electronics Engineer. Microsoft drives to Las Vegas. (2004). Connected Concept Cars. Retrieved from http://www.windowsfordevices.com/ news/NS5155857065.html Microsoft.com. (2009). Windows Automotive. Retrieved from http://www.microsoft.com/windowsautomotive/wa5/ default.mspx. Minciardi, R., & Gaetani, F. (2001). A decentralized optimal control scheme for route guidance in urban road networks. Paper presented at the Proceeding of IEEE Intelligent Transportation Systems, Oakland, CA. Mobile taiwan applications promotion project m-taiwan. Retrieved from http://www.pwlan.org.tw/mp.asp?mp=3 Monk, D., Mladenovic, D., & Skaw, M. (2003). Accelerometers for automotive applications. In J. Marek, H.P. Trah, Y. Suzuki,& I. Yokomori, (Ed.), Sensors application Vol. 4 - Sensors for automotive technology (pp. 269-296). Berlin: Wiley-VCH. 389 Compilation of References Morbe, M., & Zwiener, G. (2003). Whell-speed sensors. In J. Marek, H.P. Trah, Y. Suzuki,& I. Yokomori, (Ed.), Sensors application Vol. 4 - Sensors for automotive technology (pp. 403-415). Berlin: Wiley-VCH. MOSQUITO Project. (2005). Mosquito Architecture. White Paper. Retrieved September 20, 2009 from http://www. mosquito-online.org/ Mosra, S. R., Shanker, S., & Mahmud, S. M. (2004). An intelligent architecture for issuing intersection collision warnings (pp. 176-183). National Defense Industries Association (NDIA). OSEK VEX Portal, (n.d.). Retrieved from http://www.osek-vdx.org Muffat, M., Green, P., & Crosnier, S. (1991). European cooperation on dual mode route guidance—Perspectives for advanced research partners. Vehicle Navigation and Information Systems Conference. PaPaGO SDK. (n.d.). Retrieved from http://www.papagosdk.com/ Munson, J., Lee, S. W., Lee, D. R., Wood, D., Thompson, G., & Cole, A. (2005). A Rule-based System for Sense-andRespond Telematics Services. Workshop on End-to-end, Sense-and-Respond Systems, Applications and Services, (pp. 31-36). Musolesi, M., Hailes, S., & Mascolo, C. (2005). Adaptive routing for intermittently connected mobile ad hoc networks. In IEEE WoWMoM 2005 (pp. 183–189). Myoung, K., Heo, J., Kwon, W. H., & Kim, D. S. (2005, July). Design and implementation of home network control protocol on OSGi for home automation. In . Proceedings of the IEEE International Conference on Advanced Communication Technology, 2, 1163–1168. Nadeem, T., Shankarand, P., & Iftode, L. (2006, July). A comparative study of data dissemination models for vanets. Paper presented at the Proceeding of 3rd Annual International Conference on Mobile and Ubiquitous Systems (MOBIQUITOUS), San Jose, CA. Nagatani, T. (2002). The physics of traffic jams. Reports on Progress in Physics, 65(9), 1331–1386. doi:10.1088/00344885/65/9/203 National Highway Institute (2001). Using the National ITS Architecture for Deployment. Naumov, V., & Gross, T. (2007). Connectivity-Aware Routing (CAR) in Vehicular Ad Hoc Networks. IEEE 390 International Conference on Computer Communications (INFOCOM’07), (pp. 1919-1927). Naumov, V., Baumann, R., & Gross, T. (2006). An Evaluation Of Inter-Vehicle Ad Hoc Networks Based On Realistic Vehicular Traces. In ACM International Symposium on Mobile Ad Hoc Networking and Computing, (pp. 108-119). Next Generation Telematics Protocol website. (n.d.). Retrieved September 20 2009, from http://www.ngtp.org NMEA Data. (n.d.). Retrieved September 11, 2009, from http://gpsinformation.org/dale/nmea.htm. Noble, B., Yoon, J., & Liu, M. (2007, June 11-14). Surface Street Traffic Estimation. Paper presented at the Proceedings of International conference on Mobile systems, San Juan, Puerto Rico. Nolte, T., Hansson, H., & Bello, L. L. (2005). Automotive Communication – Past, Current and Future. In 10th IEEE Conference on Emerging Technologies and Factory Automation, 1, 985-992. Normann, N., Schulze, G., & Keller, W. (2003). Tire-pressure sensors. In J. Marek, H.P. Trah, Y. Suzuki,& I. Yokomori, (Ed.), Sensors application Vol. 4 - Sensors for automotive technology (pp. 536-544). Berlin: Wiley-VCH. Nowey, K. P. T., & Mletzko, C. (2006). Towards a Security Architecture for Vehicular ad hoc Networks. In First International Conference on Availability, Reliability and Security, (pp. 374-381). NTCIP Standard 1103 (2005). TS 3.2-1996 NTCIP Simple Transportation Management Framework - Amendment 1. NTCIP Standard 9001. (2002). The NTCIP guide. O’Donnell, M. (2003). OSEK/VDX: Driving the Standard for Optimized Embedded Systems. Embedded Computing Design. Open Service Gateway Initiative. (2003). OSGi Specification Release 3. Retrieved September 20 2009, from http:// www.osgi.org OSEK Group (2004). OSEK Implementation Language, Version 2.5. OSEK Group (2004). OSEK/VDX Network Management Concept and Application Programming Interface, Version 2.5.3. Compilation of References OSEK Group (2005). OSEK/VDX Operation System, Version 2.2.3. Ott, J., & Kutscher, D. (2006). Bundling the web: Http over dtn. In Proceeding of WNEPT2006. P. simulation VISSIM. (2009). http://www.english.ptv.de/. Palazzi, C. E., Roccetti, M., Pau, G., & Gerla, M. (2007). Online Games on Wheels: Fast Game Event Delivery in Vehicular Ad-hoc Networks. In IEEE Intelligent Transportation Systems Society, (pp. 42–49). Panichpapibo, S., & Pattara-atikom, W. (2008, Oct.). Evaluation of a neighbor-based vehicle density estimation scheme. Paper presented at the ITST 2008. 8th International Conference, Phuket. Paret, D. (2007). Multiplexed networks for embedded systems – CAN, LIN, FlexRay, Safe-by-Wire. Hoboken, NJ: Wiley. Park, C. K., Kang, J. M., Choi, M. J., Hong, J. W., Lim, Y. H., & Choi, M. S. (2008, Jan.). An Integrated Network Management System for Multi-Vendor Power Line Communication Networks. In Proceedings of the IEEE International Conference on Information Networking, (pp.23-25). Park, C. K., Kang, J. M., Choi, M. J., Hong, J. W., Lim, Y. H., Ju, S., & Choi, M. S. (2008). Definition of common PLC MIB and design of MIB Mapper for multi-vendor PLC network management. In Proceedings of the IEEE International Symposium on Power Line Communications and Its Applications, (pp.152-157). Park, J. G., Ahn, C. W., Cho, H. N., Byun, I. S., Desmons, F., & Kim, S. W. (2006). A method for representing and transporting CIM operations using binary XML in the WBEM architecture. In Proceedings of the 8th International Conference on Advanced Communication Technology, (Vol. 3, pp.20-22). Park, J. S., Lee, U., Oh, S. Y., Gerla, M., & Lun, D. (2006). Emergency Related Video Streaming in VANETs using Network Coding. In ACM International Workshop on Vehicular Ad Hoc Networks (VANET), (pp. 102-103). Pavlis, Y., & Papageorgiou, M. (1999). Simple Decentralized Feedback Strategies for Route Guidance in Traffic Networks. Transportation Science, 33(3), 264–278. doi:10.1287/ trsc.33.3.264 Pazul. K. (1999). An introduction to the CAN protocol that discusses the basics and key features. Microchip Application Note #AN713. Perkins, C. E., & Royer, E. M. (1999). Ad-hoc On-demand distance vector routing. In . Proceedings of WMCSA, 99, 90–100. Perkins, D., & McGinnis, E. (1997). Understanding SNMP MIBS. Upper Saddle River, NJ: Prentice Hall, Inc. Phung, P. H., & Sands, D. (2008). Security Policy Enforcement in the OSGi Framework Using Aspect-Oriented Programming. In Proceedings of the IEEE International Conference on Computer Software and Applications, (pp.1076-1082). Popescu, D., & Selisteanu, D. (2008). Web based Telematics Application for Robotics. In the 3rd International Multi-Conference on Computing in the Global Information Technology, (pp. 19-24). Potnis, N., & Mahajan, A. (2006) “Mobility models for vehicular ad hoc network simulations,” in proc. of the 44th annual Southeast regional conference, ACM (pp. 746–747). Prosyst website. (n.d.). Retrieved September 20 2009, from http://www.prosyst.com/ Punnoose, J., Tseng, S., Wang, S., Nikitin, V., & Schlesinger, T. E. (2001). Communications Resources Management for Advanced Telematics Applications. Transportation Society Conference. Rajkumar, R., Gagliardi, M., & Sha, L. (1995). The RealTime Publisher/Subscriber Inter-Process Communication Model for Distributed Real-Time Systems: Design and Implementation. Real-Time Technology and Applications Symposium, (pp. 66–75). Ramadas, M., Burleigh, S., & Farrell, S. (2007). Licklider transmission protocol - specification. Internet-draft. Rapport, T. S. (1996). Wireless communications principles and practice. Upper Saddle River, NJ: Prentice Hall. Raya, M., & Hubaux, J. P. (2005). The Security of Vehicular Ad Hoc Networks. In Proceedings of the 3rd ACM workshop on Security of ad hoc and sensor networks, (pp.11-21). Raychaudhuri, D., Seskar, I., Ott, M., Ganu, S., Ramachandran, K., Kremo, H., et al. (2005). Overview of the orbit 391 Compilation of References radio grid testbed for evaluation of next-generation wireless network protocols. In Proceedings of the IEEE Wireless Communications and Networking Conference, New Orleans, LA, USA. Real-Time Traffic Network Systems of Cities and Counties in Taiwan. (n.d.). Retrieved from http://www.e-traffic. com.tw Reilly, D., & Taleb-Bendiab, A. (2002). A Service-Based Architecture for In-Vehicle Telematics Systems. In the 22nd International Conference on Distributed Computing Systems Workshops, (pp.741-742). Research and Innovative Technology Administration (RITA) & U.S. Department of Transportation. (US DOT). (2002). National ITS architecture. Retrieved from http://itsarch. iteris.com/itsarch/ Rice, J., & Cho, Y. (2006). Estimating Velocity Fields on a Freeway From Low-Resolution Videos. IEEE Transactions on Intelligent Transportation Systems, 7(4), 463–469. doi:10.1109/TITS.2006.883934 Riegel, J., Wiedenmann, H., & Neumann, H. (2003). Chemical sensors for oxygen detection and air/fuel ratio contol. In J. Marek, H.P. Trah, Y. Suzuki,& I. Yokomori, (Ed.), Sensors application Vol. 4 - Sensors for automotive technology (pp. 480-499). Berlin: Wiley-VCH. Robert Bosch Gmb, H. (2006). Safety, comfort and convenience systems. Hoboken, NJ: Wiley. Robert Bosch Gmb, H. Stuttgart, Germany, (1991). CAN Specification ver. 2.0. Robert, H., & Ilya, P. (1979). A Two-Fluid Approach to Town Traffic. Science, 204(4389), 148–151. doi:10.1126/ science.204.4389.148 Robinson, C. L., Caveney, D., Laberteaux, K., & Caminiti, L. (2006, September 29). Efficient Coordination and Transmission of Data for Cooperative Vehicular Safety Applications. Paper presented at the Proceedings of the 3rd international workshop on Vehicular ad hoc networks, Los Angeles, California, USA. Rodoplu, V., & Park, M. K. (2005). An energy-efficient mac protocol for underwater wireless acoustic networks. In Proceedings of the MTS/IEEE OCEANS’05 Conference. 392 Roessler, B., & Bruns, C. (2006, Mar.). Replication of Local Road Networks for Regional Information Dissemination in VANETS. Paper presented at the Proceedings of the 3rd Workshop on positioning, Navigation And Communication (WPNC’06), University of Hannover, Germany. Rose, M. T. (1990). The Open Book: A Practical Perspective on OSI. Upper Saddle River, NJ: Prentice Hall, Inc. Rosen, J., Schulzrinne, H., Camarillo, G., Johnston, A., Peterson, J., Sparks, R., et al. (2002). SIP: Session Initiation Protocol. IETF RFC 3261. Royer, E. M., Melliar-Smithy, P. M., & Mosery, L. E. (2001) “An analysis of the optimum node density for ad hoc mobile networks,” in Proc. of the IEEE International Conference on Communications, Jun. 2001. Ryu, B., Yeung, G., Krishnan, H., Yin, J., ElBatt, T., Habermas, S., et al. (2004). Performance Evaluation of Safety Applications over DSRC Vehicular Ad Hoc Networks. Paper presented at the Proceedings of the 1st ACM international workshop on Vehicular ad hoc networks, Philadelphia, PA. S. S. of Urban Mobility. (2009). http://sumo.sourceforge. net/. Safa, H., Artail, H., & Shibli, R. (2009). An Interoperability Model for Supporting Reliability and Power-Efficient Routing in MANETs. [IJAHUC]. International Journal of Ad Hoc and Ubiquitous Computing, 4(2), 71–83. doi:10.1504/ IJAHUC.2009.023898 Saha, A. K., & Johnson, D. B. (2004) “Modeling mobility for vehicular ad hoc networks,” in Proc. of the 2nd ACM International Workshop on Vehicular Ad Hoc Networks (VANET). [Online]. Available: http://www.cs.rice.edu/ amsaha/ Resume/vanet2004/vanet2004.pdf Saha, A. K., & Johnson, D. B. (2004). Modeling mobility for vehicular ad hoc networks. ACM International Workshop on Vehicular Ad Hoc Networks (VANET), (pp. 91-92). Saito, T., Tomoda, I., Takabatake, Y., Arni, J., & Teramoto, K. (2000). Home gateway architecture and its implementation. IEEE Transactions on Consumer Electronics, 46(4), 1161–1166. doi:10.1109/30.920474 Sawamura, T., Tanaka, K., Atajanov, M., Matsumoto, N., & Yoshida, N. (2008). Adaptive Router Promotion and Group Forming in Ad-Hoc Networks. [IJAHUC]. International Compilation of References Journal of Ad Hoc and Ubiquitous Computing, 3(4), 217–223. doi:10.1504/IJAHUC.2008.018907 Schatz, O. (2003). Yaw-rate sensors. In J. Marek, H.P. Trah, Y. Suzuki,& I. Yokomori, (Ed.), Sensors application Vol. 4 - Sensors for automotive technology (pp. 297-313). Berlin: Wiley-VCH. Schiller, J., & Voisard, A. (2004). Location-Based Services. San Francisco: Morgan Kaufmann. Schilling, K. (2001). Virtual Laboratories for Mobile Robots. IFAC Workshop on Internet Based Control Education, (pp. 115-120). Schmitt, D. (2003). Chemical sensors for emission control. In J. Marek, H.P. Trah, Y. Suzuki,& I. Yokomori, (Ed.), Sensors application Vol. 4 - Sensors for automotive technology (pp. 500-508). Berlin: Wiley-VCH. Schmitt, E. J., & Jula, H. (2006, Sep.). Vehicle Route Guidance Systems: classification and Comparison. Paper presented at the Proceedings of the IEEE Intelligent Transportation Systems Conference (ITSC2006), Toronto. Scott, K. (2005). Disruption tolerant networking proxies for on-the-move tactical networks. In IEEE MILCOM 2005, (Vol. 5, pp. 3226–3231). Scott, K., & Burleigh, S. (2007). Bundle protocol specification. Internet- Draft. Segui, J., & Jennings, E. (2006). Delay tolerant networking v bundle protocol simulation. In Second IEEE International Conference on Space Mission Challenges for Information Technology. Seo, S. H., Moon, T. Y., & Kim, J. H. (2007). Smart Vehicle Management System by using Gateway, Hand-set and VMP. International Conference on Control, Automation and Systems, (pp.17-20). Seongmoon, K., & Lewis, M. E., & III, C. C. W. (2005). Optimal vehicle routing with real-time traffic information. IEEE Transactions on Intelligent Transportation Systems, 6(2), 178–188. doi:10.1109/TITS.2005.848362 Seth, A., & Keshav, S. (2005). Practical security for disconnected nodes. In Proceeding of Workshop on Secure Network Protocols (pp. 31–36). Sichitiu, M. L., & Kihl, M. (2008). Inter-Vehicle Communication Systems: A Survey. IEEE Communications Surveys & Tutorials, 10(2), 88–105. doi:10.1109/ COMST.2008.4564481 SimulationP. M. T. (2009). http://www.paramics-online. com/. SimulatorO. (2009). http://www.opnet.com/. SimulatorQ. N. (2009). http://www.scalable-networks. com/. Singh, M., Ott, M., Seskar, I., & Kamat, P. (2005). Orbit measurements framework and library (oml): Motivations, design,implementation, and features. In Proceedings of IEEE Tridentcom 2005, Trento, Italy. Skordylis, A., & Trigoni, N. (2008). Delay-Bounded Routing in Vehicular Ad-Hoc Networks. ACM international symposium on Mobile ad hoc networking and computing, (pp. 341-350). Sormani, D., Turconi, G., Costa, P., Frey, D., Migliavacca, M., & Mottola, L. (2006). Towards Lightweight Information Dissemination in Inter-Vehicular Networks. In Proceedings of the ACM International Workshop on Vehicular Inter-Networking. South Pointing Chariot. (n.d.). Retrieved September 11, 2009, from http://en.wikipedia.org/wiki/South_Pointing_Chariot. Spyropoulos, T., Psounis, K., & Raghavendra, C. S. (2004). Single-copy routing in intermittently connected mobile networks. In 2004 First Annual IEEE Communications Society Conference on Sensor and Ad Hoc Communications and Networks, (pp. 235 – 244). Spyropoulos, T., Psounis, K., & Raghavendra, C. S. (2005). Spray and wait: An efficient routing scheme for intermittently connected mobile networks. In Proceeding of the 2005 ACM SIGCOMM workshop on Delay-tolerant networking WDTN ’05, (pp. 252–259). Srisuresh, P., Kuthan, J., Rosenberg, J., Molitor, A., & Rayhan, A. (2002). Middlebox Communication Architecture and Framework. IETF Internet Draft, retrieved September 20 2009 from http://www.jdrosen.net/papers/rfc3303.txt Stallings, W. (1993). SNMP, SNMPv2 and CMIP The Practical Guide to Network Management Standards. Reading, MA: Addison-Wesley Publishing Company, Inc. 393 Compilation of References Stallings, W. (1996). SNMP, SNMPv2 and RMON. Reading, MA: Addision-Wesley Publishing Company, Inc. Steedman, D. (1993). ASN.1: The Tutorial & Reference. London: Technology Appraisals Ltd. Stepanov, I., Marron, P.-J., & Rothermel, K. (2005l) “Mobility modeling of outdoor scenarios for manets,” in Proceedings of the 38th annual Symposium on Simulation ANSS ’05, Apr. 2005 (pp. 312–322). Studer, A., Luk, M., Perrig, A, (n.d.). Efficient Mechanisms to Provide Convoy Member and Vehicle Sequence Authentication in VANETs. SUMO Simulation of Urban Mobility. (2009). Retrieved from http://sumo.sourceforge.net/ Sun Microsystems Inc. (1999, November). JINI Specification v1.0.1. Retrieved from http//:www.sun.com/JINI/ Sun Microsystems Inc. (2000). Java Media Framework API Guide, 2000. Retrieved from http://java.sun.com/products/ java-media/jmf/ Sun, W., Yamaguchi, H., Yukimasa, K., & Kusumoto, S. (2006). GVGrid: A QoS Routing Protocol for Vehicular Ad Hoc Networks. IEEE International Workshop on Quality of Service, (pp. 130-139). Sun, Y., Huang, W. L., Tang, S. M., Qiao, X., & Wang, F. Y. (2007). Design of an OSEK/VDX and OSGi-based embedded software platform for vehicular applications. In Proceedings of the IEEE International Conference on Vehicular Electronic and Safety, (pp. 1-6). Sun, Y., Wang, F. Y., Wang, Z. X., Qiao, X., & Wang, K. F. (2006). A scheduling algorithm for vehicular application specific embedded operating systems. In Proceedings of IEEE International Conference on Systems, Man and Cybernetics, (pp. 2535-2540). Sven Jaap, M. B., & Wolf, L. (2005) “Evaluation of routing protocols for vehicular ad hoc networks in city traffic scenarios,” in in Proc. of the 5th International Conference on Intelligent Transportation Systems Telecommunications (ITST). Symington, S., Farrell, S., Weiss, H., & Lovell, P. (2007). Bundle security protocol specification. Internet-draft. Symington, S., Farrell, S., Weiss, H., & Lovell, P. (2007). Delay-tolerant networking security overview. Internetdraft. T. N. S. ns 2.(2009). http://www.isi.edu/nsnam/ns/index. html. Taleb, T., Sakhaee, E., Jamalipour, A., Hashimoto, K., Kato, N., & Nemoto, Y. (2007). A Stable Routing Protocol to Support ITS Services in VANET Networks. IEEE Transactions on Vehicular Technology, 56(6), 3337–3347. doi:10.1109/ TVT.2007.906873 Tan, K., Zhang, Q., & Zhu, W. (2003). Shortest path routing in partially connected ad hoc networks. In Proceedings of IEEE Globecom 2003, (Vol. 2, pp. 1038–1042). Tariq, M. M. B., Ammar, M., & Zegura, E. (2006). Message ferry route design for sparse ad hoc networks with mobile nodes. In Proceedings of the seventh ACM international symposium on Mobile ad hoc networking and computing MobiHoc ’06 (pp. 37–48). Thangavelul, A., & Bhuvaneswari, K. Kumar’, K., SenthilKumarl, K., & Sivanandam, S.N., (2007). Location Identification and Vehicle Tracking using VANET (VETRAC). In IEEE International Conference on Signal Processing, Communications and Networking 2007. The Network Simulator. ns2.(2009). Retrieved from http:// www.isi.edu/nsnam/ns/ The zebranet wildlife tracker. (2009). Retrieved from http:// www.princeton.edu/ mrm/zebranet.html The, I. E. E. E. 802.15.4 WPAN Group. (2009). Retrieved from http://www.ieee802.org/15/pub/TG4.html The, O. N. WORLD. (2005). Retrieved from http://www. onworld.com Third Generation Partnership Project [3GPP]. (2009). Retrieved from http://www.3gpp.org Thompson, J. P. (1998). Web-based enterprise management architecture. IEEE Communications Magazine, 36(3), 80–86. doi:10.1109/35.663331 Tier. (2009). Retrieved from http://tier.cs.berkeley.edu/ wiki/Home TIGER (Topologically Integrated GEographic Encoding and Referencing). (2009). http://www.census.gov/geo/ www/tiger/. 394 Compilation of References Tomkewitsch, R. v. (1991). Dynamic route guidance and interactive transport management with ALI-SCOUT. Journal of IEEE Transactions on Vehicular Technology, 40(1), 45–50. doi:10.1109/25.69971 Tonguz, O. K., Wisitpongphan, N., Parikh, J. S., Bai, F., Mudalige, P., & Sadekar, V. K. (2006). On the Broadcast Storm Problem in Ad hoc Wireless Networks. International Conference on Broadband Communications, Networks and Systems, (pp. 1-11). Tonguz, O., Wisitpongphan, N., Bai, F., Mudalige, P., & Sadekar, V. (2007). Broadcasting in VANET. Mobile Networking for Vehicular Environments, (pp. 7-12). Torge, W. (1991). Geodesy. Berlin: Walter de Gruyter. Traffic Message Channel. (n.d.). Retrieved September 11, 2009, from http://en.wikipedia.org/wiki/Traffic_Message_Channel. TRANSIMS. http://transims.tsasa.lanl.gov/. Transportation Research Board. (2004). Transportation In An Aging Society: A Decade of Experience. Transportation Research Board. Treiber, M., Hennecke, A., & Helbing, D. (2000) “Congested traffic states in empirical observations and microscopic simulations,” IPhys. Rev., vol. 62, no. 2. Tropos networks. (2009). Retrieved from http://www. tropos.com Tsai, H. M., Tonguz, O. K., Saraydar, C., Talty, T., Ames, M., & Macdonald, A. (2007). ZigBee-based intra-car wireless sensor networks: A case study. IEEE Wireless Communications, (pp. 68-77). Tseng, Y. C., Chen, J. J., & Cheng, Y. L. (2007). Design and Implementation of a SIP-based Mobile and Vehicular Wireless Network with PUSH Mechanism. IEEE Transactions on Vehicular Technology, 56(6), 3408–3420. doi:10.1109/ TVT.2007.906986 Tseng, Y. C., Ni, S. Y., Chen, Y. S., & Sheu, J. P. (2002). The Broadcast Storm Problem in a Mobile Ad Hoc Network. [WINET]. ACM Wireless Networks, 8(2), 153–167. doi:10.1023/A:1013763825347 Unwired wireless. (2009). Retrieved from http://www. unwired.com.au/ UPnP Forum. (2009). Retrieved from http://www.upnp. org Vahdat, V., & Becker, D. (2000). Epidemic routing for partially connected ad hoc networks. Technical Report CS-200006, Duke University. Vahid, F., & Givargis, T. (2002). Embedded system design: A unified hardware/software introduction. Hoboken, NJ: Wiley. Wan, S., Tang, J., & Wolff, R. S. (2008). Reliable Routing for Roadside to Vehicle Communications in Rural Areas. In IEEE International Conference on Communications, (pp. 3017-3021). Wang, C., Sohraby, K., Jana, R., Ji, L., & Deneshmand, M. (2008). Voice communications over ZigBee networks. IEEE Communications Magazine, 121–127. doi:10.1109/ MCOM.2008.4427240 Wang, F. Y., Li, J. X., & Liu, X. J. (2004). OSEK/VDX and OSGi-based Embedded Platform for Vehicular Software and its Key Techniques. Chinese High Technology Letters, (pp.131-136). Weber, S., Cahill, V., Clarke, S., & Haahr, M. (2003). Wireless ad hoc network for dublin: A large-scale ad hoc network test-bed. ERCIM News, 54. Retrieved from http://www. ercim.org/publication/ErcimNews/enw54/weber.html Weiss, E., Gehlen, G., Lukas, S., & Rokitansky, C. (2006). MYCAREVENT- Vehicular Communication Gateway for Car Maintenance and Remote Diagnosis. In Proceedings of the IEEE International Conference on Computers and Comunication, (pp. 318-323). WiChorus, Inc. (2009). Retrieved from http://www.mustafaergen.com/asn_gateway.pdf Wills, J., Ye, W., & Heidemann, J. (2006). Low-power acoustic modem for dense underwater sensor networks. In Proceedings of the First ACM International Workshop on UnderWater Networks (WUWNet), Los Angeles, CA, (pp. 79–85). New York: ACM. Retrieved from http://www.isi. edu/ohnh/PAPERS/Wills06a.html WirelessCar website (n.d.). Retrieved September 20 2009, from http://www.wirelesscar.com/ Wisitpongphan, N., Bai, F., Mudalige, P., Sadekar, V., & Tonguz, O. (2007). Routing in sparse vehicular ad hoc wireless 395 Compilation of References networks. IEEE Journal on Selected Areas in Communications, 25(8), 1538–1556. doi:10.1109/JSAC.2007.071005 Wisitpongphan, N., Tonguz, O. K., Parikh, J. S., Mudalige, P., Bai, F., & Sadekar, V. (2007). Broadcast Strom Mitigation Techniques in Vehicular Ad Hoc Networks. IEEE Wireless Communications, 14(6), 84–94. doi:10.1109/ MWC.2007.4407231 Wood, L., Eddy, W. M., Ivancic, W., McKim, J., & Jackson, C. (2007). Saratoga: a delay-tolerant networking convergence layer with efficient link utilization. In Proceeding of IWSSC’07. Wood, L., McKim, J., Eddy, W., Ivancic, W., & Jackson, C. (2008). Saratoga: a convergence layer for delay-tolerant networking. Internet draft. Wood, L., Peoples, C., Parr, G., Scotney, B., & Moore, A. (2007). Tcps protocol radius:the distance where timers prevent communication. In Proceeding of IWSSC’07. Woodrow, B., & Thomas, A. (1998). Dingus Human Factors in Intelligent Transportation Systems. Mahwah, NJ: Lawrence Erlbaum Associates. World Geodetic System. (n.d.). Retrieved September 11, 2009, from http://en.wikipedia.org/wiki/World_Geodetic_System WTP1. 0 Specification (2004). Electronics and Telecommunications Research Institute (ETRI), Korea. Wu, H., Fujimoto, R., Guensler, R., & Hunter, M. (2004, Oct.). Mddv: a mobility-centric data dissemination algorithm for vehicular networks. Paper presented at the Proceedings of ACM Vehicular ad hoc networks (VANET), Philadelphia, PA. Wu, H., Fujimoto, R., Hunter, M., & Guensler, R. (2005). An architecture study of infrastructure-based vehicular networks. In ACM MSWiM, Montreal, Canada, (pp.36-39). Wunderlich, K. E., Kaufman, D. E., & Smith, R. L. (2000). Link travel time prediction for decentralized route guidance architectures. Journal of IEEE Transaction on Intelligent Transportation Systems, 1(1), 4–14. doi:10.1109/6979.869017 WWWOFFLE. (2009). Retrieved from http://www.gedanken.demon.co.uk/wwwoffle/ 396 Hudson Valley Transportation Management Center (HVTMC) (n.d.). Retrieved from http://www.hudsonvalleytraveler. com/Glossary.html Open Service Gateway Initiative (OSGi). (2009). Retrieved from http://www.osgi.org WiMAX Forum. (2007). WiMAX End-to-End Network Systems Architecture Stage 3: Detailed Protocols and Procedures, Release 1.1.0. WiMAX Forum. (n.d.). Wimax deployment consider for fixed wireless Access in 2.5GHz and 3.5GHz licensed bands. Xiang, W., Richardson, P., & Guo, J. (2006). Introduction and preliminary experimental results of wireless access for vehicular environments (WAVE) systems. In Proceedings of 3rd Annual International Conference on Mobile and Ubiquitous Systems, (pp.1-8). Xu, Q., Mak, T., Ko, J., & Sengupta, R. (2004). Vehicle-toVehicle Safety Messaging in DSRC. In ACM International Workshop on Vehicular Ad Hoc Networks (VANET), (pp. 19-28). Xu, Y. N., Kim, Y. E., Cho, K. J., Chung, J. G., & Lim, M. S. (2008). Implementation of FLexRay communication controller protocol with application to a robot system. In Proceedings of 15ht International Conference on Electronics, Circuits and Systems, (pp. 994-997). Yamashita, T., Izumi, K., & Kurumatani, K. (2004). Car navigation route information sharing for improvement of traffic efficiency. Paper presented at the Proceeding of 7th international IEEE conference on Intelligent Transportation Systems, Washington, D.C., USA. Yang, H., Luo, H., Ye, F., Lu, S., & Zhang, L. (2004). Security in mobile ad hoc networks: Challenges and solutions. IEEE Wireless communications, 11(1), 38-47. Yang, P., & Chuah, M. C. (2006). Context-aware multicast routing scheme for disruption tolerant networks. In Proceedings of ACM international workshop on PE-WASUN ’06 (pp. 66–73). Ye, Q., Cheng, L., Chuah, M. C., & Davison, B. D. (2006). Os-multicast: On-demand situation-aware multicasting in disruption tolerant networks. In . Proceedings of IEEE VTC, 1, 96–100. Compilation of References Yokomori, I., & Suzuki, Y. (2003). Pressure sensors. In J. Marek, H.P. Trah, Y. Suzuki,& I. Yokomori, (Ed.), Sensors application Vol. 4 - Sensors for automotive technology (pp. 314-332). Berlin: Wiley-VCH. Zhao, J., & Cao, G. (2006). VADD: Vehicle-Assisted Data Delivery in Vehicular Ad Hoc Networks. IEEE Computer Communications, (pp. 1-12). Yoon, H., Kim, J., Tan, F., & Hsieh, R. (2008). On-demand Video Streaming in Mobile Opportunistic Networks. In IEEE International Conference on Pervasive Computing and Communications (PerCom), (pp. 80–89). Zhao, W., Ammar, M., & Zegura, E. (2004). A message ferrying approach for data delivery in sparse mobile ad hoc networks. In Proceedings of the 5th ACM international symposium on Mobile ad hoc networking and computing MobiHoc ’04 (pp. 187–198). Yousefi, S., Mousavi, M. S., & Fathy, M. (2006). Vehicular Ad Hoc Networks (VANETs): Challenges and Perspectives. International conference on ITS Telecommunications, (pp. 761-766). Zhao, W., Ammar, M., & Zegura, E. (2005). Controlling the mobility of multiple data transport ferries in delaytolerant network. In . Proceedings of IEEE INFOCOM, 2, 1407–1418. Yu, F., & Biswas, S. (2007). Self-Configuring TDMA Protocols for Enhancing Vehicle Safety with DSRC Based Vehicle-to-Vehicle Communications. IEEE Journal on Selected Areas in Communications, 25(8), 1526–1537. doi:10.1109/JSAC.2007.071004 Zhao, W., Ammar, M., & Zegura, E. (2005). Multicasting in delay tolerant networks: Semantic models and routing algorithms. In Proceeding of Sigcomm Workshop in DTN, (pp. 268–275). Zeng, X., Bagrodia, R., & Gerla, M. (1998). Glomosim: A library for the parallel simulation of large-scale wireless networks. In The 12th Workshop on Parallel and distributed Simulation. PADS’98, Banff, Alberta, Canada (pp. 154–161). Zhan, F. B. (1997). Three fastest shortest path algorithms on real road network: data structure and procedures. Journal of geographic information and decision analysis, 1(1), 69-82. Zhang, D., Xiao, H. W., & Hackbarth, K. (2004). OSGi based service infrastructure for context aware automotive telematics. In . Proceedings of IEEE International Conference on Vehicular Technology, 5, 2957–2961. Zhang, Y., Zhao, J., & Cao, G. (2007). On Scheduling Vehicle-Roadside Data Access. In Proceedings of the ACM International Workshop on Vehicular Inter-Networking. Zhang, Z., & Fei, Z. (2007). Route design for multiple ferries in delay tolerant networks. In Proceedings of IEEE Wireless Communications and Networking Conference (pp. 3460–3465). Zhao, Y. (1997). Vehicle Location and Navigation Systems. Boston: Artech House. Zheng, N. N., Tang, S., Cheng, H., Li, Q., Lai, G., & Wang, F. W. (2004). Toward Intelligent Driver-Assistance and Safety Warning Systems. IEEE Intelligent Systems, 19(2), 8–11. doi:10.1109/MIS.2004.1274904 Zheng, Q., Hong, X., & Ray, S. (2004). Recent advances in mobility modeling for mobile ad hoc network research. In Proceedings of the 42nd ACM annual Southeast regional conference (ACM-SE), (70-75). New York: ACM Press. Zhou, Y., Wang, X., & Zhou, M. (2006). The Research and Realization for Passenger Car CAN Bus. In Proceedings of the IEEE International Forum on Strategic Technology, (pp. 244-247). Zigbee Alliance. (2006). ZigBee specification: ZigBee Document 053474r13. Zimmermann, H.-M., Gruber, I., & Roman, C. (2005) “A voronoi-based mobility model for urban environments,” in in Proc. Of the European Wireless 2005 (EW05) . Zonoozi, M. M., & Dassanayake, P. (1997). User mobility modeling and characterization of mobility patterns . IEEE Journal on Selected Areas in Communications, 15(7), 1239–1252. doi:10.1109/49.622908 397 398 About the Contributors Chung-Ming Huang received the B.S. degree in electrical engineering from National Taiwan University on 1984/6, and the M.S. and Ph.D. degrees in computer and information science from The Ohio State University on 1988/12 and 1991/6. Currently, he is a Distinguished Professor of Dept. of Computer Science and Information Engineering, National Cheng Kung University, Taiwan, R.O.C. He also serves as (i) Director of the Promotion Center for the Telematics Consortium (PCTC), Ministry of Education (MOE), Taiwan, R.O.C. and (ii) Principal Project Reviewer of Industrial Development Bureau and Department of Industrial Technology, Ministry of Economic Affairs (MOEA), Taiwan, R.O.C. He has published more than 200 referred journal and conference papers in wireless and mobile communication protocols, interactive multimedia systems, audio and video streaming and formal modeling of communication protocols. His research interests include wireless and mobile network protocol design and analysis, media processing and streaming, web technologies, and network applications and services. Yuh-Shyan Chen received the B.S. degree in Computer Science from Tamkang University, Taiwan, R. O. C., in June 1988 and the M.S. and Ph.D. degrees in Computer Science and Information Engineering from the National Central University, Taiwan, R. O. C., in June 1991 and January 1996, respectively. He joined the faculty of Department of Computer Science and Information Engineering at Chung-Hua University, Taiwan, R. O. C., as an associate professor in February 1996. He joined the Department of Statistic, National Taipei University in August 2000, and joined the Department of Computer Science and Information Engineering, National Chung Cheng University in August 2002. Since 2006, he has been a Professor at the Department of Computer Science and Information Engineering, National Taipei University, Taiwan. Prof. Chen is now serving as chair of Institute of Communication Engineering, National Taipei University, Taiwan, ROC, and Vice Chair of Task Force on “Telecommunications” of Intelligent Systems Applications Technical Committee, IEEE Computational Intelligence Society from 2007. Prof. Chen served as Editor-in-Chief of International Journal of Ad Hoc and Ubiquitous Computing (SCIE), Editorial Board of Telecommunication System Journal (SCIE), EURASIP Journal on Wireless Communications and Networking (SCIE), and Mobile Information Systems (SCIE). He served as Guest Ed