Trust model for TRIDNT trust based routing Protocol

Abstract

In a mobile ad-hoc network, nodes cannot rely on any fixed infrastructure for routing purposes. Rather, they have to cooperate to achieve this objective. However, performing network functions consumes energy and other resources. Therefore, some network nodes may decide against cooperating with others; selfish nodes. Also security issues are more paramount in such networks even more so than in wired networks. In particular these networks are extremely under threat to insider; malicious nodes; especially through packet dropping attacks. Selfish and malicious nodes are termed as misbehaving nodes. Giving the selfish nodes the incentive to cooperate, while isolating the misbehaving nodes have been an active research area recently. In this paper, we introduce our Trust Model for TRIDNT (Two node-disjoint Routes scheme for Isolating Dropper Node in MANET) the reactive trust based routing Protocol, which allows some degree of node selfishness to give an incentive to the selfish nodes to declare its selfishness behavior to its neighbors, which reduce the searching time of misbehaving nodes to search for the malicious nodes only. In the proposed TRIDNT trust model, the trust among nodes is represented by trust score, which consist of direct and indirect trust, with addition of new term called the cooperation score, the trust calculation is based on the beta probability density function. TRIDNT use both DLL-ACK and end- to-end TCP-ACK as monitoring tools to monitor the behavior of routing path nodes: if the data packet successfully transmitted, then the path nodes trust value are updated positively; otherwise, if a malicious behavior is detected the path searching tool starts to identify the malicious nodes and update its trust value negatively and isolate them from the routing path and the network. We use the cooperation score to measure the unselfishness behavior of a suspect node and control the allowed selfishness behavior. Finally we calculate an accurate trust value threshold to distinguish between trustworthy or untrustworthy node.