Exploiting Mobility for Trust Propagation in Mobile Ad Hoc Networks

Trust plays an important role in protecting the security of mobile ad hoc networks. The node mobility brings challenge to trust propagation, where traditional graph-based propagation methods are difficult to apply. In this paper, we propose a trust establishment and propagation scheme by exploiting natural mobility of mobile ad hoc network nodes. We estimate the moving state of neighboring nodes, and select the group of nodes with higher movement diversity as the next hop. In this way, we expedite trust propagation with the help of mobility instead of treating it as a hurdle. Two classes of mobility models are considered: random direction model and cluster-based model. Extensive experiments are conducted on trust propagation performance using our scheme. Results show that the detection rate of highly untrustworthy node is improved by about 400% in random direction mobility model and about 300% in cluster-based mobility model compared to static case. Further, we observe that trust convergence time and communication overhead varies dramatically in different mobility models. Based on these observations, we discuss the usage of different mobility models and metrics in different trust applications in mobile ad hoc networks.