A Three-Factor Based Trust Model for Anonymous Bacon Message in VANETs
Song Zhang, Rui He, Yunpeng Xiao, Yanbing Liu
Abstract
The periodic beacon message is one of the most common messages in vehicular ad hoc networks (VANETs) and plays an important role in enhancing road safety. To protect the identity and location privacy of the vehicle, the beacon messages should be anonymous and unlinkable under a conditional privacy-preserving authentication protocol. However, due to malicious intent or sensor failure, even certified vehicles may share dishonest anonymous beacon messages, thereby endangering road safety. In this paper, according to the three important trust factors including expectation, risk and confidence, we propose a three-factor based data-centric trust model (ERC-TM) to judge the trustworthiness of anonymous beacon messages. First, we evaluate the expectation value, risk value and confidence value of a beacon message respectively. Then, according to the qualitative relationship between trust and the three factors, we propose a trust decision based on the three factors to judge whether the beacon message is trustworthy. The proposed trust model makes the trusted beacon messages have the characteristics of high expectation, low risk and high confidence. We define different attacks according to the modification degree of beacon messages, and conduct simulation experiments in two common traffic scenarios. Simulation results show that the proposed model can effectively identify malicious beacon messages. Furthermore, we analyze the impact of different attacks and the confidence values of the proposed model. In addition, the impact of each trust factor is also analyzed through ablation experiments, and the relevant requirements are discussed.