Cluster and Recommendation Based Multi-granularity Trust Model in P2P Network

A trust model for P2P network is proposed in this thesis, namely CRM, a cluster and recommendation based multi-granularity trust model. Compared to the current trust model, some problems can be solved better, for instance, these disadvantages contain great costs due to iterative calculations, and security such as disguise, slandering caused by anonymous and open characteristics of P2P networks. The simulation results show the CRM model is more robust in resisting evil behaviors efficiently. Therefore the quality of service and the success rate for network deals in P2P network are highly improved.