Reputation-Based Trust Model and Anti-attack Mechanism in P2P Networks

In the security community, one main focus has been devoted to the enhancement of access control approaches with new authentication and authorization capabilities with the server to restrict others´ ability. However, in the peer-to-peer (P2P) networks, no such centralized server exists, especially for the pure P2P, and the peers have to interact with unfamiliar peers. To prevent or discourage peers from distributing invalid or malicious content into the network, a reputation-based mechanism with direct and indirect feedbacks is presented in this paper. Some solutions within the mechanism for malicious attacks such as colluding, whitewashing and behavior changing are also investigated.