TrueTrust: a feedback-based trust management model without filtering feedbacks in P2P networks

Abstract

With the increases of P2P applications and their users, the attacks on P2P networks have also increased significantly. To handle the problem, researchers have proposed many kinds of P2P trust management models, in which the feedback-based ones took the major part. Due to the existence of false feedbacks, however, the effectiveness of such a trust management model depends heavily on how effective the feedback filtering mechanism is. This paper proposes a novel trust management model named TrueTrust, which makes use of all the feedbacks on a peer to calculate the peer’s credibility without using feedback filtering mechanism. Also, our trust management model calculates a peer’s credibility for a requester heavily relying on the requester’s past feedbacks on the peer’s services. This means that if a requester slandered a peer’s services for its collusive group, then it is difficult for the requester to receive services from the peer any more even though the services are authentic due to the lower peer’s credibility calculated for the requester. This reflects the idea of retributive justice and thus discourages rational peers from giving false feedbacks. The simulation results indicate that our trust management model could both curb malicious attacks and improve successful download rate in P2P networks.