Research On QoS-Driven Incentive Machanism In P2P Environment

Peer-to-Peer Network is the typical representative of current open Network, has drawn tremendous attention from both business and academia. However, peers in P2P Network show the characteristics of rationality, self-interest etc., and try to maximize their utilities without considering the effects of their behaviors on QoS utility. Due to the lack of incentive and restriction, we cannot make ensure that peers abide by the intention of system design, thus result in the existence of noncooperation, free-riding, and then.seriously affect network's QoS performance. Therefore, this thesis possesses a great academic and practical value for study how to establish an effective mechanism to restrain peers'behavior and ensure the QoS performance of P2P networks.In order to overcome the above problems, this thesis will study the incentive mechanisms to analyze the factors that influence on the QoS. We aim at reducing the maintenance overhead in the process of file sharing, as well ensuring the network's performance through improving the strategies of sharing. The main contents and contribution of this thesis includs the following aspects:·We proposed an incentive model with supervision mechanism based on game theory, aim to promote fair distribution of resources, and reduce the maintenance overhead caused by the changes of topology, as well improve the resource utilization of system.·We combined service time and resources contribution to measure the peer's comprehensive contirbution, and analyzed the effection of weight factors on peer's incentive value, to reflect the behavior and profit of peers.·We reconstructed the system architecture based on P2PSim, as well imporved and optimized the traditional DHT-based Chord system. Then we verified the effectiveness of proposed mechanism through the simulation, to provide a feasible solution for designing P2P incentive mechanisms.