An Attack-resistant Trust Inference Model for Securing Routing in Vehicular Ad-hoc Networks

The problem of routing security in vehicular ad hoc networks has become a major concern for researchers. Compared with cryptography-based solutions, trust-based solutions are considered to be more acceptable as a promising approach, which mainly define two operations: trust computing and security application. In this paper we first study trust properties and construct a novel trust inference model, where two trust attributes named subjective trust and recommendation trust, are selected to quantify the trust level of a specific vehicle. The SCGM(1,1)-weighted Markov prediction algorithm based on fluctuation recognition is utilized to calculate the subjective trust accurately, meanwhile a new evaluation method of recommendation credibility based on the feedback mechanism is introduced for the calculation of recommendation trust. We subsequently exploit the advantages of mesh structure and design a light-weight trust-aware multicast routing protocol (i.e., LTMRP), which can establish secure and reliable communication paths by selecting trusted relay vehicles. To further improve the routing efficiency, two mechanisms including forwarding nodes reuse mechanism and trust-aware route handoff mechanism are proposed. Simulation verification and comparison of the relevant trust models and trust-based protocols are accomplished by three simulation platforms (i.e., Netlogo, SUMO, and NS-2).The convincing results show that our trust model obtains strong resistibility to multiple malicious attacks. Moreover, examined in four routing metrics, the experimental results indicate that LTMRP performs better than the other relevant protocols.