车辆自组网以其可延展的多跳网络模式成为一种典型的车联网结构，而基于身份的管理模式是当前车辆管理模式在车联网系统中的自然承继。项目研究车辆自组网中基于身份的安全机制，针对当前VANET基于身份管理的研究中存在的RSU认证负荷过大、车辆通信关系泄露隐私信息以及车辆合法身份滥用等三个方面的问题，研究基于身份的车辆系统安全机制，包括：车辆的代理批认证技术，以车辆代理其它车辆认证缓解RSU的认证负荷；车辆轨迹无关的隐私保护技术，以分布式秘密共享的群组认证和RSU认证结合的方式，消除RSU认证带来的位置隐私泄露问题；多个车辆合法身份合谋攻击检测方法，将嫌疑节点聚类为团组结构以团组整体行为来判定合谋攻击。此外，项目还研究车辆中嫌疑节点及嫌疑群组发展态势预测及该态势对整个网络系统安全影响的评估方法，并基于该影响建立系统的激励与惩罚机制以实现对态势发展的有效遏制。

Vehicular ad hoc network (VANET) is receiving extensive attention for its scalable structure of multi-hop communication model. VANET manages vehicles based on car identity in accordance with the pattern in the real word. Thus identity management and authentication are the specific key issues of security in VANET. In this project, identity-based security mechanism will be studied. And some problems, such as massive authentication burden on RSU, privacy leakage from communication relationship, and the abuse of legal identity will be discussed and solved. To establish a secure identification-based VANET, we will research and obtain the corresponding key technologies for each issue. Firstly, Proxy Vehicle Batch Verification Alogrithm, a vehicle is chosen as a proxy to authenticate other vehicles in order to remit the heavy authentication burden on RSU; Secondly, Trajectory-irrelevant Privacy Preserving Technology, it integrates RSU authentication and group authentication which is based on distributed secret sharing, to eliminate the privacy disclosure caused by RSU authentication; Thirdly, Multi-identity Collusion Attack Detection Method, susceptible nodes are assembled as a clique and collusion attack is measured based on the whole clique instead of a node. In addition, the evaluation methods for the development trend of the susceptible cliques will be studied. And based on the evaluated results, stimulation and punishing mechanism will be studied to restrain the susceptible trend. In aspects of the aforementioned four issues, our proceed work has obtained some good results which also show the feasibility of our research approach of this proposal.