Context sensitive trust based geographic opportunistic routing in mobile ad hoc networks

摘要

Position based opportunistic routing (POR) is a stateless, robust, and reliable geographic routing protocol in Mobile AdHoc NETwork (MANET). The opportunistic routing embraces broadcast property of wireless channels and utilizes it for opportunistic forwarding. Both the malicious node behavior and the backup nodes' behavior are equally treated as malicious in the existing misbehavior detection mechanisms. Hence, incorporating a general trust model in POR is not combative with routing attacks. It is necessary to determine whether the misbehavior is likely a result of malicious activity or due to the backup scenario of opportunistic forwarding. On the other hand, if context-sensitive trust information is available on every node, it ensures a fair decision making and also supports secured routing in an opportunistic approach. This work investigates the utilization of context attributes along with generic trust model to allow POR for secure and reliable data forwarding. This paper introduces context-sensitive trust for choosing the data forwarding node in POR (CPOR) to assist opportunistic routing in selecting the trusted optimal data forwarding node and to cope with both security and reliability of communications. The proposed work exercises both coarse- and fine-grained trust evaluation to strengthen the trustworthiness. The coarse-grained trust measure includes positive progress per hop and behavioral attribute of the nodes in terms of routing service. The fine-grained trust evaluation differs the opportunistic routing environment from the adverse scenarios and aids the source node such that it builds a highly trusted positive progress set using contextual attributes. The fine-grained trust evaluation deduces the ideal contextual information such as the link quality, battery energy, and the backup service to determine the accurate trust value of nodes. As a result, it involves optimal routes and enables CPOR to maintain the routing performance equal to the POR even in the presence of malicious nodes in the network. The simulation results demonstrate that the packet delivery ratio of the proposed context-aware trust model is substantially high even when 50% of the total nodes are found malicious and outperforms the trust-aware opportunistic routing protocol (TAOR).