Design And Analysis Of P2P Content Distribution Technology Based On Cooperative Game Theory

P2P content distribution network is the production of combining the advantages ofP2P network and content distribution technique, which is a hot topic in networkapplication researching currently. However, there are a lot of selfish peers in the actualnetwork. They occupy a large number of bandwidth resource but not willing to provideresources and service for others, which produce “Free-riding” and “Tragedy ofcommons”. It is contrary to the original design concept and severely affects the overallperformance of the system. For this problem, we introduce the cooperative game theoryof economics. And we use three techniques including Shapley value, NBS (NashBargaining Solution) and Repeated Game in cooperative game theory to solve theincentive problem of peer cooperation to complete the sharing resource for two typicalP2P content distribution network.Firstly, we summarize the P2P content distribution network, including the researchbackground, current status of the study and research significance. We focus on thecurrent incentive mechanism in the P2P content network. Then, P2P technique, contentdistribution network and game theory are introduced. Finally, the design and analysis ofP2P content distribution technology based on cooperative game theory is proposed forthe unsolved problems. The goal is suppress the free-riding behavior and improve theoverall performance of the P2P content distribution network.This paper mainly include three aspects: first, Shapley value based storagemanagement in P2P video-on-demand systems. Shapley value concept is employed inthis scheme to encourage cooperation among peers by allocating the profit of cooperationin a fair way. Such a cooperative game theory model based on storage managementscheme can significantly improve the data hitting ratio in a P2P VOD system. Second,the social P2P network is a typical distribution network, in which peers with socialattribute can easy to cooperate for sharing bandwidth. But the problems of optimal strategy and fairness need to be solved. So a NBS-based bandwidth allocation scheme forsocial P2P network is proposed. The problem is formulated as a Nash bargaining game,and the Nash bargaining solution for a two-peer bargaining game is derived with theanalysis of Pareto optimality and fairness. Third, a scheme about repeated game theorybased penalty-incentive for social P2P network is presented. Due to every peer is selfishand we can’t guarantee that all peers are always faithful during the entire process ofcooperation in social P2P network. In order to prevent the occurrence of peers’ fraudbehavior, we design a named Punishment-Forgiving-Restoring (PFR) mechanism. Theproposed mechanism aims at effectively suppressing the occurrence of peers’ fraudbehavior by the conditional punishments, promoting the honest cooperation among socialpeers, and improving the average payoffs of the whole network.The analytical and experimental results show that the proposed mechanisms notonly effectively prohibit free-riding behavior, but also complete the rational allocation ofresource, and significantly improve the efficiency and fairness of data distribution andget more average profit comparing with the existing incentive mechanism. In addition, itis significant to enhance the overall performance and promote the further development ofP2P content distribution applications.