Vehicular Ad Hoc Network Security and Privacy
Description
More details
Other editions
Additional editions
Persons
Content
List of Figures xi
List of Tables xv
Acronyms xvii
Preface xix
1 INTRODUCTION 1
1.1 Background 1
1.2 DSRC AND VANET 2
1.2.1 DSRC 2
1.2.2 VANET 3
1.2.3 Characteristics of VANET 6
1.3 Security and Privacy Threats 7
1.4 Security and Privacy Requirements 8
1.5 Challenges and Prospects 9
1.5.1 Conditional Privacy Preservation in VANETs 9
1.5.2 Authentication with Efficient Revocation in VANETs 10
1.6 Standardization and Related Activities 11
1.7 Security Primitives 13
1.8 Outline of the Book 17
References 17
2 GSIS: GROUP SIGNATURE AND ID-BASED SIGNATURE-BASED SECURE AND PRIVACY-PRESERVING PROTOCOL 21
2.1 Introduction 21
2.2 Preliminaries and Background 23
2.2.1 Group Signature 23
2.2.2 Bilinear Pairing and ID-Based Cryptography 23
2.2.3 Threat Model 23
2.2.4 Desired Requirements 24
2.3 Proposed Secure and Privacy-Preserving Protocol 25
2.3.1 Problem Formulation 25
2.3.2 System Setup 27
2.3.3 Security Protocol between OBUs 29
2.3.4 Security Protocol between RSUs and OBUs 38
2.4 Performance Evaluation 41
2.4.1 Impact of Traffic Load 43
2.4.2 Impact of Cryptographic Signature Verification Delay 43
2.4.3 Membership Revocation and Tracing Efficiency 45
2.5 Concluding Remarks 47
References 47
3 ECPP: EFFICIENT CONDITIONAL PRIVACY PRESERVATION PROTOCOL 51
3.1 Introduction 51
3.2 System Model and Problem Formulation 52
3.2.1 System Model 52
3.2.2 Design Objectives 54
3.3 Proposed ECPP Protocol 55
3.3.1 System Initialization 55
3.3.2 OBU Short-Time Anonymous Key Generation 56
3.3.3 OBU Safety Message Sending 62
3.3.4 OBU Fast Tracking Algorithm 63
3.4 Analysis on Conditional Privacy Preservation 64
3.5 Performance Analysis 66
3.5.1 OBU Storage Overhead 66
3.5.2 OBU Computation Overhead on Verification 66
3.5.3 TA Computation Complexity on OBU Tracking 68
3.6 Concluding Remarks 69
References 69
4 PSEUDONYM-CHANGING STRATEGY FOR LOCATION PRIVACY 71
4.1 Introduction 71
4.2 Problem Definition 73
4.2.1 Network Model 73
4.2.2 Threat Model 74
4.2.3 Location Privacy Requirements 75
4.3 Proposed PCS Strategy for Location Privacy 75
4.3.1 KPSD Model for PCS Strategy 75
4.3.2 Anonymity Set Analysis for Achieved Location Privacy 79
4.3.3 Feasibility Analysis of PCS Strategy 85
4.4 Performance Evaluation 86
4.5 Concluding Remarks 89
References 89
5 RSU-AIDED MESSAGE AUTHENTICATION 91
5.1 Introduction 91
5.2 System Model and Preliminaries 93
5.2.1 System Model 93
5.2.2 Assumption 93
5.2.3 Problem Statement 94
5.2.4 Security Objectives 95
5.3 Proposed RSU-Aided Message Authentication Scheme 96
5.3.1 Overview 96
5.3.2 Mutual Authentication and Key Agreement between RSUs and Vehicles 96
5.3.3 Hash Aggregation 98
5.3.4 Verification 99
5.3.5 Privacy Enhancement 100
5.4 Performance Evaluation 101
5.4.1 Message Loss Ratio 102
5.4.2 Message Delay 102
5.4.3 Communication Overhead 104
5.5 Security Analysis 105
5.6 Concluding Remarks 106
References 107
6 TESLA-BASED BROADCAST AUTHENTICATION 109
6.1 Introduction 109
6.2 Timed Efficient and Secure Vehicular Communication Scheme 110
6.2.1 Preliminaries 110
6.2.2 System Formulation 112
6.2.3 Proposed TSVC Scheme 113
6.2.4 Enhanced TSVC with Nonrepudiation 118
6.2.5 Discussion 123
6.3 Security Analysis 129
6.4 Performance Evaluation 129
6.4.1 Impact of Vehicle Moving Speed 131
6.4.2 Impact of Vehicle Density 132
6.5 Concluding Remarks 134
References 134
7 DISTRIBUTED COOPERATIVE MESSAGE AUTHENTICATION 137
7.1 Introduction 137
7.2 Problem Formulation 138
7.2.1 Network Model 138
7.2.2 Security Model 139
7.3 Basic Cooperative Authentication Scheme 140
7.4 Secure Cooperative Authentication Scheme 141
7.4.1 Evidence and Token for Fairness 142
7.4.2 Authentication Proof 145
7.4.3 Flows of Proposed Scheme 146
7.5 Security Analysis 147
7.5.1 Linkability Attack 147
7.5.2 Free-Riding Attack without Authentication Efforts 147
7.5.3 Free-Riding Attack with Fake Authentication Efforts 148
7.6 Performance Evaluation 148
7.6.1 Simulation Settings 148
7.6.2 Simulation Results 149
7.7 Concluding Remarks 150
References 151
8 CONTEXT-AWARE COOPERATIVE AUTHENTICATION 153
8.1 Introduction 153
8.2 Message Trustworthiness in VANETs 156
8.3 System Model and Design Goal 159
8.3.1 Network Model 159
8.3.2 Attack Model 159
8.3.3 Design Goals 160
8.4 Preliminaries 160
8.4.1 Pairing Technique 160
8.4.2 Aggregate Signature and Batch Verification 160
8.5 Proposed AEMAT Scheme 161
8.5.1 System Setup 161
8.5.2 Registration 162
8.5.3 SER Generation and Broadcasting 162
8.5.4 SER Opportunistic Forwarding 162
8.5.5 SER Aggregated Authentication 163
8.5.6 SER Aggregated Trustworthiness 165
8.6 Security Discussion 168
8.6.1 Collusion Attacks 168
8.6.2 Privacy Protection of Witnesses 168
8.7 Performance Evaluation 169
8.7.1 Transmission Cost 169
8.7.2 Computational Cost 169
8.8 Concluding Remarks 170
References 170
9 FAST HANDOVER AUTHENTICATION BASED ON MOBILITY PREDICTION 173
9.1 Introduction 173
9.2 Vehicular Network Architecture 175
9.3 Proposed Fast Handover Authentication Scheme Based on Mobility Prediction 176
9.3.1 Multilayer Perceptron Classifier 176
9.3.2 Proposed Authentication Scheme 178
9.4 Security Analysis 183
9.4.1 Replay Attack 183
9.4.2 Forward Secrecy 183
9.5 Performance Evaluation 184
9.6 Concluding Remarks 185
References 186
Index 187
Preface
Road safety has been drawing increasing public attention, and there has been extensive effort in both industry and academia to mitigate the impact of traffic accidents. Recent advances in wireless technology bring promising new ways to facilitate road safety and traffic management, in which each vehicle, equipped with wireless communication devices [referred to as onboard units (OBUs)], is allowed to communicate with vehicles, other as well as with roadside units (RSUs), which are located at critical sections of the road, such as traffic lights and stop signs.
With OBUs and RSUs, a self-organized network called a vehicular ad hoc network (VANET) can be formed. Recently, this has emerged as a promising approach toward increasing road safety and efficiency, as well as improving driving experience. These goals will be accomplished through a wide variety of vehicle applications enabled by communication between vehicles, such as emergency braking warning. While society experiences tremendous benefits from adopting the new technologies, we also continue to face challenges; the biggest challenge is always how to address security and privacy issues that may be caused by the adoption of a new technology. The attractive features of VANETs will inevitably incur higher risks for abuse, if we do not take security and privacy issues into consideration before the wide deployment of such networks.
Being a special implementation of wireless ad hoc networks or mobile ad hoc networks (MANETs), a VANET has many unique features and applications. First, the connectivity among nodes (vehicles and RSUs) can often be highly transient and a one-time event; two vehicles may remain within their transmission ranges, or within a few wireless hops, for only a very limited period of time. As a result, vehicular network topology is highly dynamic. Further, a VANET is a huge network, which can potentially consist of millions of nodes (on-road vehicles and RSUs). Such size makes it very challenging to guarantee security and privacy in VANETs, particularly regarding message authenticity and integrity, as well as protecting user-related privacy information, such as the driver's name and the car's license plate, model, and traveling route. Unfortunately, existing studies on (and solutions for) communication security and privacy preservation cannot work effectively in VANETs, since they do not take the scalability and communication overhead into consideration. Message authentication is a common tool for ensuring information reliability, but it faces a challenge in VANETs. Particularly when a vehicle receives a large number of messages, traditional authentication mechanisms may generate unaffordable computational overhead on the vehicles, and bring unacceptable delay to time-critical applications (e.g., accident warning). Another challenge is the privacy concerns of vehicular communication, where the identity, position, and movement track of a specific vehicle should not be obtained by an unauthorized third party. We will refer to the combined concepts of message authenticity and privacy as anonymous message authentication.
In this book, we focus on message authentication and privacy issues in VANETs. We first provide an overview of security and privacy issues in VANETs, as well as the challenges facing VANETs in addressing these issues.
Chapter 2 identifies the unique security and privacy requirements of communications between different types of communication devices, including OBUs and RSUs in VANETs. We determined the most suitable cryptographic primitives and designed a secure and privacy-preserving protocol, which utilizes a combination of group signature and identity (ID)-based signature techniques to addresses these unique security and privacy requirements for vehicular communications.
Chapter 3 further exploits the unique challenges in privacy-preserving VANETs, i.e., how to efficiently deal with the growing revocation list while achieving conditional traceability. Based on the on-the-fly short-term anonymous key generation between OBUs and RSUs, we proposed an efficient conditional privacy preservation protocol, which is characterized by providing the conditional privacy preservation, improving efficiency in terms of the minimized anonymous key storage at each OBU and fast verification on safety messages.
Chapter 4 discusses the pseudonym changing strategy for location privacy in VANETs, as even though an OBU holds a large number of pseudonyms in VANETs, if the pseudonym does not change at the right time and right place, location privacy could still be violated. To enable vehicles to achieve high-level location privacy, we proposed an efficient pseudonym changing at social spots (PCS) strategy, where the social spots are the places where many vehicles temporarily gather.
Afterward, we take a cooperative approach toward addressing the technology's challenges of complex anonymous message authentication. Cooperation on anonymous message authentication in VANETs can be defined as vehicles and RSUs working together to ensure the integrity of messages received by each individually, as well as verifying that messages are indeed from legitimate users. Cooperation can take many forms and occur in many ways in VANETs, for example, either between vehicles and RSUs, or only among vehicles. Cooperation can also occur in many different ways, based on the roles of vehicles and RSUs in groups. For example, the resource-rich RSUs are usually seen as trusted entities in VANETs, since RSUs are usually deployed by governments or service providers, and their locations are fixed. As a result, a straightforward approach for message authentication in VANETs is to leverage the vast resources of RSUs and take advantage of their fixed locations. RSUs can be used to assist vehicles to authenticate messages received by the vehicles, largely by allowing resource-rich and trustworthy RSUs take the lead processing role in message authentication. In the case of cooperation only among vehicles, which will be very common in the early stages of VANET adoption (due to a lack of RSUs), each vehicle can probabilistically validate a certain percentage of its received messages in accordance with its own computing capacity, and report any invalid message detected. When work all units together, redundant effort in message authentication and verification can be minimized, if not entirely eliminated. Further, cooperation can occur by taking into consideration the context of messages transmitted over the vehicular networks. For example, of all vehicle communication network applications, dissemination of emergency messages to the vehicles in a specific area is one of the most crucial. The fast propagation of emergency and local warning messages to the approaching vehicles will be helpful for preventing secondary accidents, especially in conditions where visibility is impeded, such as fog. In most cases, a VANET performs such an emergency message propagation in a multihop transmission manner, particularly in the suburban areas where fewer RSUs are installed. Given any emergency, it is expected that multiple sensing vehicles in the area could detect the same common event, and therefore, taking advantage of this property to cross-validate the emergency event could possibly serve as a promising approach toward enhancing the overall security level of VANETs. Such a method of cross-checking the emergency event by collecting the feedback of witnesses is defined as a voting mechanism, which was originally used to detect the misbehaving nodes in a distributed ad hoc network without any centralized security authority. This kind of cooperation is often applied to deal with special types of messages, such as emergency messages, and the mechanism can be migrated to VANETs to enhance the overall security of emergency events authentication.
We classify cooperative authentication mechanisms in VANETs into four categories: RSU-aided authentication (Chapter 5), TESLA-based authentication (Chapter 6), distributed cooperative authentication (Chapter 7), and context-aware cooperative authentication (Chapter 8). For each category, we introduce a corresponding protocol for message authentication, and will also analyze security, efficiency, and effectiveness of these proposed cooperative authentication protocols. Both theoretical analysis and simulation results show that cooperative authentication is a promising and effective way to achieve secure message authentication for vehicular communications.
Because of the movement of the vehicles, the vehicles can roam among RSUs deployed along the roadsides. The final chapter looks into the challenges in realizing seamless mobility in VANETs. By considering some intrinsic features of vehicular communication networking, such as predictable vehicle movement, we introduce a seamless authentication scheme based on mobility prediction to achieve fast authentication and reduce the authentication delay.
The book primarily presents our research results of anonymous message authentication in VANETs, but also provides a comprehensive survey of existing challenges and solutions in security and privacy in VANETs.
We wish to thank many people whose insightful comments and suggestions have helped us significantly improve our research work. In particular, we would like to acknowledge the following researchers who have collaborated with us on this exciting research topic described in the book: Prof. Xuemin (Sherman) Shen, Prof. Pin-Han Ho, Prof. Haojin Zhu, Dr. Chenxi Zhang, Xiaoting Sun, Dr. Xiaoyu Wang, Dr. Xiaohui Liang, Dr. Tom H. Luan, and Dr. Xu Li. Our discussions and collaboration with them provide a critical foundation for the current book. Also,...
