车载自组网的集成统一信任性理论与技术研究

Before executing tasks in Vehicular Ad-hoc Networks (VANETs), vehicles have to meet three preconditions: trusted identity (authentication process), trusted state (verification process), and trusted behavior (routing process). From the perspectives of the shared attributes among vehicles, resources and environments, the objective of this project is to explore the correlation and constraint between attribute system and trust system, and then to provide an integrated and unified new theory and technology of vehicular trust including identity, state and behavior. First, according to correlation between attributes and trust, we build a subject, object and environment attribute system of VANETs, and then focus on classification, symbolization, policy modeling, and the other related aspects. Second, we combine authentication model based on attributes and threshold cryptography, and meanwhile according to the local mobility characteristics, we design a novel threshold authentication model to realize trust of identity. Third, we apply trust computing technology in allocating tasks of evaluation centre to some vehicles which would launch verification process together on those vehicles with trusted identity. A level of trust will be determined by attribute integrity to realize trust of state. Finally, those vehicles with trusted identity and trusted state are allowed to join in routing process. We propose the concept of attribute similarity, and design a subjective trusted routing model with consideration to reliability and length of path to realize trust of behavior. The realization of trust of vehicular identity, state and behavior could provide trust guarantee to application services.

车辆欲加入车载自组网执行任务有三个前提：可信身份（认证过程）、可信状态（评估过程）和可信行为（路由过程）。本项目旨在以车辆、资源和环境具有的属性为主线，探索属性系统与信任系统间的关联和约束关系,研究建立车辆间集成统一的身份、状态和行为信任性的新理论与新技术。首先，依据与信任性的关联程度，建立车载自组网的主体、客体和环境属性系统，研究属性的分类、符号化和策略模型等；其次，结合基于属性的认证模型和门限密码，并根据车辆的局部移动特点，设计门限认证模型，实现身份信任；再次，利用可信计算技术将评估中心的任务分散给若干车辆承担，对身份信任的车辆发起评估过程，根据车辆属性完整性程度赋予信任等级，实现状态信任；最后，具有可信身份和状态的车辆便可参与路由过程,提出属性相似度概念和计算方法，设计兼顾可靠性和长度的主观信任路由模型，实现行为信任。实现车辆身份、状态和行为的信任性将为后续的应用服务提供信任性保障。

车载自组网是由一组带有中短距离无线收发装置的车辆组成的一类特殊的移动自组网，为驾驶员提供超视距诸如碰撞避免、区域告警和信息分发等各类应用服务，具有可观的经济价值、深远的社会影响和重要的环保意义。车辆欲加入车载自组网执行任务有三个前提:可信身份(认证过程)、可信状态(评估过程)和可信行为(路由过程)。本项目首先依据与信任性的关联程度,建立车载自组网的主体、客体和环境属性系统,研究属性的分类、符号化和策略模型等;其次,结合基于属性的认证模型和门限密码,并根据车辆的局部移动特点,设计门限认证模型,实现身份信任;再次,利用可信计算技术将评估中心的任务分散给若干车辆承担,对身份信任的车辆发起评估过程,根据车辆属性完整性程度赋予信任等级,实现状态信任;最后,具有可信身份和状态的车辆便可参与路由过程,提出属性相似度概念和计算方法,设计兼顾可靠性和长度的主观信任路由模型,实现行为信任。重要结果分别为：提出了建立身份和状态信任性的SAV4AV协议、适用于车载自组网的情景感知密钥交换协议VIKE、分析了蠕虫病毒在静态和动态交通流下的传播动力学行为、建模了车载协议栈网络层路由行为和物理层信号传播特性，分析了互联车辆的交通拥塞传播动力学行为。本项目实现的车辆身份、状态和行为信任性将为后续的应用服务提供信任性保障，为车载自组网提供相对独立完整的信任性解决方案。