PeerMate: A malicious peer detection algorithm for P2P systems based on MSPCA

Abstract

Many reputation management schemes have been proposed to assist peers in choosing the most trustworthy collaborators in a P2P environment where honest peers coexist with malicious ones. While these schemes indeed generally provide some useful information regarding the reliability of peers, they still suffer from various attacks such as slandering, collusion, etc. Consequently, being able to detect the malicious peers plays a critical role in the successful functioning of these mechanisms, and this is our focus in this paper. First, we divide the malicious peers into several categories. Second, we introduce PeerMate, a malicious peer detection algorithm based on Multiscale Principal Component Analysis and Quality of Reconstruction, to detect malicious peers in Reputation-based P2P systems. Finally, we experimentally demonstrate that PeerMate is able to detect malicious peers accurately and efficiently.