Enhancing the reliability of Ad-hoc networks through fuzzy trust evaluation

Ideally, in Ad-hoc networks all comprising nodes act cooperatively in establishing the communication paths from various sources to destinations. However, in practice such an ideal situation may not always be realized. Some nodes may act selfishly, refusing to accept the burden of forwarding packets. Some other nodes may act maliciously. Either way, such nodes degrade the network reliability or even disrupt its operation altogether. To alleviate the problems caused by the presence of such nodes, this paper proposes to establish quantifiable trust levels for the comprising nodes and to use them in making routing decisions. The proposed approach uses fuzzy logic concepts in evaluating the trust levels of the nodes. The trust levels are then used in the routing process. Using OPNET and MATLAB, the proposed approach is validated and further analyzed. It is shown that the utilization of this approach can result in significant enhancements of the performance and reliability of Adhoc networks.