Design And Effectiveness Of Incentive Mechanisms For Peer Cooperation In P2P Networks Based On Networked Evolutionary Game Theory

Recently, P2P networks have been extensively applied to file sharing and media streaming services. Since peers in P2P networks are autonomous, and resource sharing mainly relies on the voluntary behavior of peers, cooperation among peers is crucial for the performance of P2P networks. Resource sharing incurs extra costs to peers, and rational peers in the network seek to maximize their profits, thus peers with free-riding behaviors will flourish in the network. The free-riding behaviors can seriously compromise the performance of the network, thus, the incentive mechanism, which is one of efficient methods to prohibit peers'selfish behaviors in autonomous networks, has begun to receive considerable attention. Game theory can model the conflict of interests between an individual and the whole network effectively, and it has been used extensively in the study of incentive mechanisms.In this thesis, we review the existing research relating to the design of incentive mechanisms and their effectiveness analysis, and on such basis, we detailedly study the design of incentive mechanisms for peer cooperation and the effectiveness analysis of designed incentive mechanisms in P2P networks. We provide an incentive mechanism based on the adjustment of peers'transaction willingness, and networked evolutionary game theory based effectiveness analysis frameworks for P2P file-sharing and media streaming networks. The main contributions of this dissertation are listed as follows:1. An incentive mechanism based on the adjustment of transaction willingness of peers is proposed. Existing researches on the incentive mechanisms adopt the reciprocity mechanism based on historical transactions, while neglecting the transaction willingness of peers. These incentive mechanisms lead to high costs on storage and processing of information. Therefore, we provide a transaction willingness adjustment mechanism that follows the rule of "optimism when winning and cautiousness when losing", which considers the transaction willingness of peers and the influence of learning behaviors in transactions on the transaction willingness. The effectiveness of the incentive mechanism is verified based on networked evolutionary game theory. Simulation results and analysis show that the proposed adjustment mechanism of transaction willingness can promote peers'cooperative behaviors in resource sharing, and cooperation can be guaranteed more effectively when the adjustment sensitivity of peers'transaction willingness is higher.2. An effectiveness analysis framework for incentive mechanisms in P2P file-sharing networks is proposed. Since previous researches do not consider the influence of peers' relationships in transactions on the effectiveness of incentive mechanisms, we use peers' transaction overlay networks to model peers'relationships. In addition to the influence of peers'irrational behaviors on the effectiveness of incentive mechanisms, we also incorporate the influence of peers'leaving and rejoining behaviors. The effectiveness of incentive mechanisms is studied in scenarios where distributions of benefits of peers'sharing services are homogeneous and heterogeneous. To verify the effectiveness of the proposed framework, the generosity based incentive mechanism is analyzed within the proposed framework. Simulation results and analysis demonstrate that the framework can give the effectiveness condition and the reason why the incentive mechanism can promote cooperation among peers.3. A multi-player asymmetric media streaming sharing game model and performance analysis frameworks for incentive mechanisms in P2P media streaming networks are proposed. Previous researches related to the effectiveness analysis of incentive mechanisms in P2P media streaming networks mainly consider two scenarios. In the first scenario, each peer selects its allocation strategy of the shared bandwidth independently. In the second scenario, each peer allocates its shared bandwidth based on a specific mechanism. We propose effectiveness analysis frameworks for reciprocity incentive mechanisms in these two scenarios respectively. To verify the effectiveness of proposed frameworks, the proportional allocation based incentive mechanism is analyzed within the proposed frameworks. Simulation results and analysis show that our frameworks can give the effectiveness condition of the incentive mechanism and the reason why the incentive mechanism can promote cooperation among peers.