CuboidTrust: a global reputation-based trust model in peer-to-peer networks

Peer-to-Peer communication model has the potential to harness huge amounts of resources. However, some recent studies indicate that most of current Peer-to-Peer systems suffer from inauthentic resource attacks. One way to cope with these attacks is to constitute a reputation-based trust model to help evaluating the trust values of peers and predicting their future behaviors. In this paper, we propose a global reputation-based trust model, called CuboidTrust. It builds four relations among three trust factors including contribution, trustworthiness and quality of resource, and applies power iteration to compute the global trust value of each peer. The experimental results show that CuboidTrust performs efficiently, and significantly decreases the count of inauthentic resource downloads under various threat models.