Authors: Yujie Song; Yue Cao; ChakLam Cheong; Debiao He; Kim-Kwang Raymond Choo

 

Title: CAT: A Consensus-Adaptive Trust Management based on the Group Decision Making in IoVs

 

Journal:  IEEE Transactions on Information Forensics and Security

 

Abstract: Securing Internet of Vehicles (IoVs) systems against common threats such as false message injection remains challenging, and one typical approach is to deploy trust management solutions. In this work, we propose a Consensus-Adaptive Trust management (CAT) based on the Group Decision Making (GDM) in IoVs. Specifically, in our approach the consensus levels are calculated to measure the difference of opinions (trust values) among vehicles. To estimate the reliability of consensus levels, the divergence between consensus levels is calculated, namely: consensus level similarity. GDM allows us to dynamically adjust the opinion of vehicles (namely: Consensus Reaching Process, CRP). Then, a credit guarantee mechanism is designed to improve the efficiency of CRP and seek out malicious vehicles quickly. To empower the adaptability of trust management for changing environments in IoVs, CAT dynamically manages opinions of vehicles, self-confidence, and their consensus thresholds according to the feedback of delivered messages. Extensive simulation results show the potential of CAT operating in high-risk scenarios, and outperforming other competing baseline methods in terms of accuracy, precision, recall, and F-score.