A Novel Trust Evaluation Model Based on Grey Clustering Theory for Routing Networks

The trust solutions to routing networks are faced with the evaluation of behavior trust and how to exactly evaluate the behaviors under the circumstance of existing recommend deceptive behavior such as providing fake or misleading recommendation. In this paper, by learning trust relationship from routing network, a trust evaluation model based on Grey Clustering Theory is proposed. The model adopts improved Bayes theory to evaluate the behavior trust. By introducing Grey Clustering Theory, the model clusters the recommend node to different trust classes according to recommend credibility and calculates the recommend weight to resistance the fraud recommends information from the deceptive node. Simulation results show that trust evaluation model based on Grey Clustering Theory cannot only effectively evaluate the routing node behavior but also has better anti-attack performance, anti-deception performance and higher attack node detection rate.