BITA: A Behavioral Interaction Trust Analysis Framework for False Message Detection in VANETs

Abstract

Vehicle-to-Vehicle (V2V) and Vehicle-to-Infrastructure (V2I) communications are the key to intelligent transportation services and the basis of Vehicular Ad Hoc Networks (VANETs). But due to the open communication setup, VANETs can be easily subject to false message dissemination attack, which can have a significant impact on road safety and traffic management. To overcome this challenge, an analysis framework is proposed in this paper that is called Behavioral Interaction Trust Analysis (BITA) to identify false messages reliably. The model incorporates concepts on reliability assessment of the mobility services, cooperative forwarding assessment, temporal credibility adaptation, modelling of behavioral consistency and propagation of trust within the neighbourhood and yields an adaptive trust score for each vehicle. The proposed method is based on behavioral interaction analysis to detect malicious activities which is different from the conventional methods based on reputation. Experimental tests were performed with different levels of vehicle densities and attack sets. The proposed BITA framework was able to attain 96.4% accuracy, 95.4% precision, 95.3% recall, and 3.5% false positive rate. Moreover, it achieved low communication overheads in comparison to the current trust-based methods. The results gained in this paper show that BITA can be effectively used to boost trust management and false message detection performance in dynamic vehicular communication environments.