Research And Implement Of Trust Mechanism In P2P Networks Based On Reputation

In recent years, P2P technology as a distributed computing technology has made rapid development and has attracted close attention of domestic and foreign academic researchers. Hence a lot of applications based on P2P technology have emerged. P2P technology known as peer-to-peer computing mode has broken the traditional Client/Server mode which allows network nodes to share files equally rather than a server / client distinction.However, the problem of P2P network security has greatly restricted the rapid development of P2P technology. The trust mechanism as a novel security solution has been proposed as a promising way to handle security problems in P2P network environment. The basic idea of trust mechanism is to let service recipient rate service provider after each transaction, and use the aggregated ratings to derive a trust value which could help other peers to decide whether or not to transact with this service provider. Trust mechanism has become a research focus of P2P networks.Firstly, this thesis chooses the trust mechanism of P2P networks as research background, explores and summarizes the current typical trust mechanisms of P2P networks as well as their shortcomings. Secondly, this thesis proposes a reputation-based trust model for P2P networks which could find out malicious peers by detecting the fluctuation of global trust value and similarity of recommendation. Both the penalty factors and the convergence of iteration of global trust value are introduced and proved. Experiments results show that this model is superior to classic models in numerous respects. Finally, a P2P security prototype is developed based on the reputation-based trust model, and the operation results show that the trust model could effectively evaluate trust value of peers in P2P networks and has a good effect in increasing success rate of interaction.