A Computational Reputation Model in P2P Networks Based on Trust and Distrust

Reputation mechanism can be used for choosing the peers suitable to collaborate with in P2P networks, so reputation and trust management has attracted many researchers’ attentions. In this paper, a reputation management model for P2P systems is presented, in which trust and distrust are integrated, and multiple levels are differentiated on each aspect. The Confidence Index (CI) is introduced in the model for representing the confidence in evaluation results. Then a peer selection algorithm is designed for P2P collaborations. Simulation evaluation with different settings shows it effectively isolates malicious peers, and enhances the performances of the networks.