Research And Design On P2P Incentive Model Based On Game Theory

With the advancement of information technology and the development of Internet, the Peer-to-Peer system becomes larger and larger, and consequently the problem of free-riding is becoming more and more serious. Accordingly, the incentive mechanism, that incentive nodes occupy and share the resource fairly rationally, emerges as required.On the basis of analysis and summary of the existing incentive mechanism, and to overcome the problem of free-riding, a P2P incentive mechanism is proposed based on RDEC algorithm in this paper, which discourages free-riding by letting peers favor uploading to other peers, who have been proven to be good up-loaders according to its responsivity as well as upload bandwidth. The algorithm is analyzed and proved via simulation experiments. At the end of the paper, we design a system, which includes the incentive mechanism, using the Resource Distribution algorithm based on Earning and Contribution (RDEC).The main details are as follows:1. We analyze the problem of free-riding, which is more and more serious and has great impact on P2P system. Also we discuss on the importance to solving the problem of free-riding on P2P system.2. We analyze and compare the existing micro-payment model, direct reciprocity model, reputation-based mechanisms and punishment mechanism. At the same time, we discuss their advantages and disadvantages.3. To overcome the drawbacks of the incentive models, we prove that the characteristics of P2P network are in accordance with Nash Equilibrium theory through comparing Nash Equilibrium theory in the game theory with the problems in P2P network. We propose the RDEC algorithm based on incentive mechanism model using the Bertrand duopoly model of Nash Equilibrium theory. We compare the P2P system of none incentives and based on Tit-for-tat algorithm, numerical experimental results show that the three performance evaluations of the proposed RDEC algorithm, which are the selfish nodes, the overall effectiveness of the system and the payload of system, are much better than that of the non-incentive mechanism of the P2P system and that of the Tit-for-tat with incentives. The results show that the given algorithm is correct and effective. Our RDEC algorithm can incentive to the node and achieve the purpose of maximizing the system utility.4. We design and implement a prototype of the mechanism of the system using the proposed RDEC algorithm based on the incentive mechanism. In the system design, we adopt a fully distributed design and a single node on the hierarchical staicture and minimize the system interface to ensure that the given mechanism could be run in the major peer-to-peer system. Through adding the incentive mechanism, which based on RDEC algorithm, to the existing open-source system KAN8888, we compare the test results, which show that the proposed method improves the performance of the system greatly.