Modeling and analysis of PeerTrust-like trust mechanisms in P2P Networks

To counter malicious peers in P2P systems, PeerTrust-like schemes take advantage of similarity between peers to compute service provider´s trust value. To derive the similarity, there currently exist some approaches documented which however mostly are interested in algorithm or process enhancement, the analytic study in theory for these PeerTrust-like mechanisms might be missed. This paper focuses on the basic problems in PeerTrust-like trust schemes and makes the following distinctive contributions. It formalizes the generic framework of trust mechanisms in a probabilistic manner; gives the mathematical description for PeerTrust-like trust mechanisms; attempts to figure out some unclear questions: based on PeerTrust-like schemes, how to compute similarity via Euclidean distance; should all feedback be counted in; how to compare PeerTrust-like mechanisms with other distance/similarity computations.