Securing multihop vehicular message broadcast using trust sensors

Abstract

The ad hoc wireless exchange of position and velocity information between vehicles enables a plethora of new applications that can increase the safety and efficiency of driving. Efficient and reliable flooding mechanisms for vehicular applications mandate correct and timely received positions, vehicular safety applications even more so. This work first assesses the impact of different position faking attackers on the ";goodput" of Multi-Hop Vehicular Beacon Broadcast (MHVB-B), a dissemination mechanism for vehicular networks. Then we use a set of known and simple heuristics to improve the detection of fake positions within MHVB-B data and briefly assess their impact on the goodput. At the core of this work, we define a framework for integrating arbitrary trust sensors using Bayesian reasoning and describe a way to determine their contribution to the overall assessment of message trustworthiness, that we model as a conditional probability. © 2008 International Federation for Information Processing.