On trust models for communication security in vehicular ad-hoc networks

Abstract

Vehicular ad-hoc networks (VANETs) are prone to various network attacks due to its open, dynamic, and distributed nature. It is of increasing importance to assess the trustworthiness of vehicle nodes and messages for communication security in such networks. In this paper, we propose an integrated security scheme to help nodes in VANETs identify the legitimacy of messages for proper decision making. Within the proposed scheme, we first construct trust models to quantify the direct and indirect reputation of a participating node. Particularly, we use the certain-factor model to measure the node’s direct reputation by observing its communication behavior, and synthesize the information from the node’s neighbors and road side unit (RSU) reports to determine its indirect reputation. Both types of reputation are then combined using the uncertain deductive method. Based on these trust models, we further design an attribute-weighted K-means method to identify legitimate messages and a routing method for message delivery in VANETs. We also design and incorporate a dynamic game model-based incentive mechanism to encourage nodes to take cooperative actions and restrain them from selfish behaviors. Extensive simulations show that the proposed security scheme exhibits superior performances over existing methods in alleviating negative effects caused by selfish or malicious nodes.