An Anti-collusion Trust Model in P2P Networks

摘要

This paper proposed an anti-collusion trust model in order to resist collusion attack which has become the first threat in trust models of P2P networks. This trust model could detect the existing colluding peers and give them a penalty by introducing two penalty factors. When we evaluate the global trust value of a peer in the network, we consider not only the local trust values but also the recommended trust value of this peer. The convergence of iteration of global trust value is also taken into consideration and proved as well. Experiment results show that our anti-collusion trust model is effective in resisting collusion and also increasing the fraction of authentic downloads.