A security management scheme using a novel computational reputation model for wireless and mobile ad hoc networks

Robust trust and reputation evaluation services are significant to ensure the security of mobile ad hoc networks, as centralized administration is not easy to be applied and these networks do not have fixed infrastructures. However, malicious behaviors obviously can make wireless and mobile networks at risk so malicious nodes should be detected and excluded. In this paper, we study the issue of how to evaluate efficiently a node's reputation in a distributed approach. First, a set of management mechanisms is presented to prevent effectively malicious nodes from entering the trusted community. Then we formulate a comprehensive computational reputation model. Through a set of extensive simulation experiments based on the ns-2 simulator, our simulation results demonstrate that our system indeed are more sensitive to suspicious behaviors and peers, and thereby improper behaviors within a community are prevented effectively. Therefore, in dynamic but agile environments, our computational reputation system exhibits high adaptation and low complexity, so that it leads to secure communication among mobile nodes.