Resources Auction Incentive Mechanism Based On Reputation In P2P Networks

In 21st century the information century, The application on P2P hase exceeded the application on Web. And it become the topic persist focus by the researcher P2P networks become the direction of one of the most active areas of research in distributed systems There are many free riding in P2P because of the characteristics of P2P' peers such as distributed, dynamic change, anonymous, etc. As there are more and more peer-to-peer systems in the world and they become larger and larger, How to resolve the problem of fair resources sharing in peer-to-peer system is becoming more and more serious. In the early stage of peer-to-peer systems, very little attention was paid to incentive mechanism and most systems didn't contain such kind of design. As a result, more and more problems due to lack of incentive mechanism appear. So the incentive mechanism in P2P networks is the most important research in distributed systems It is a key factor of how to improve the capability of P2P, that how to encourage P2P peers of rational and self interesting to contribute peer's resources effectively and reasonably and improve the utility of P2P's resource. In this thesis, we present a novel model to encourage P2P peers of rational and self interesting to contribute peer's resources.The major research work in this thesis can be viewed as follow: 1.Constructed a model of peer's reputation base on the interactive history between peers and the application of small world network theory in psychology.The application of P2P networks usually requires cooperation among various peers to complete and Because of the distributed, dynamic change, anonymous in P2P,it is an important issue to select right peers to cooperate. But we can get a well done by establishing a reputation systems. Inspired by the theory of small worlds and strength relation of Granovetter, we compute the similarity of peers through interactive history between peers. We set the more weight on peer's reputation evaluation we want to get the reputation from weak relation to some peers than strong relation peer's. We compute a peer' the reputation in distributed by using direct information and indirect or environment information2. Constructed Incentive mechanism base on peer's resources utility.The goal of Incentive mechanism is design to a comparatively perfect mechanism. In the mechanism the resources peer's get from the systems are related to the resources peer's contribute to the systems. The early Incentive mechanism in P2P was built according to the resources which peers contributed to P2P networks, so it is a preferred idea that peers which contributed more resources can get return by higher peer's resources utility. We constructed an incentive model that the peers which contribute their resources positivly have priority to get resources from P2P networks.3. Constructed a resources pricing model in P2P .We must make a price for peers' resources because auction is the measure of resource allocation. Price of resources is related to peers' reputation, the number of peers in auction and the constraints of auction. In this thesis we do some research in auction constraints about price and fee cancellation.4. Constructed a resources auction incentive mechanism base on Reputation in P2P Networks. Auction has many advantages in economics incentive mechanism, auction taking into account the peer's private information, being rational and self interesting. We provide incentive mechanism so that peers can get benefit if they show the true value of information about their resources . Meanwhile it make peers flexible, effective, extensible, instant feedback and help peers make rational investment decisions. Peers' reputation is the key factor in auction, So we constructed a resources auction incentive mechanism based on reputation in P2P Networks.Eventually, we make simulation experiment. We analysed the evaluation of the reputation of peers' in P2P under different conditions, and the incentive mechanism of auction. The experimental results show the feasibility and efficiency of our method.