An Improved Trust Model in P2P

Based on the Bayesian network trust model, an improved trust model is proposed in this paper. It includes two parts: one is the improved searching resource peers algorithm, which can reduce the redundancy packages received by peers and enhance the reliability of system; the other is the improved trust computing and updating method, which applies logarithm for evaluation and the number of services to compute and update the trust, so that it can prevent malicious peers from attacking normal peers efficiently, meanwhile restrain malicious peers from increasing their trust respectively by collusion in its clique. The sketch maps of trust value illustrate that the new model controls the behavior of malicious peer in P2P more efficiently