Poisonedwater: an adaptive approach to reducing the reputation ranking error in P2P networks

Abstract

This paper preliminarily proposes a reputation ranking algorithm called “Poisonedwater” to resist front peer attack—peers that gain high reputation values by always cooperating with other peers and then promote their malicious friends through passing most of their reputation values to those malicious peers. Specifically, we introduce a notion of Poisoned Water (PW) that iteratively floods from identified malicious peers in the reverse direction of the incoming trust links towards other peers. Furthermore, we propose the concept of Spreading Factor (SF) that is logistically correlated to each peer's PW level. Then, we design the new reputation ranking algorithm seamlessly integrated with peers' recommendation ability (represented as SF), to infer the more accurate reputation ranking for each peer. Simulation results show that, in comparison with Eigentrust, Poisonedwater can significantly reduce the ranking error ratio up to 20%, when P2P systems exist many malicious peers and front peers.