Issue No. 06 - June (2010 vol. 21)
DOI Bookmark: http://doi.ieeecomputersociety.org/10.1109/TPDS.2009.122
Hiroshi Nishida , Oregon State University, Corvallis
Thinh Nguyen , Oregon State University, Corvallis
For many P2P systems, implementing right incentives and policies to promote efficient and fair resource sharing is the key to improve the overall system performance. In this paper, we propose a points-based incentive mechanism named Global Contribution (GC) approach that efficiently and naturally maintains fairness in a P2P network. In this approach, a proposed GC algorithm first calculates a global score for each peer that accurately reflects its bandwidth contribution to the entire network. Then, these scores are used in a proposed data transfer policy to determine whether one peer can download data from other peers. Thus, the GC approach achieves: 1) efficiently preventing free-riding, 2) naturally balancing the upload and download amounts in each peer, and 3) reducing rejections in transactions between cooperative peers. Moreover, the GC algorithm requires only private transaction history as an input and can be fully decentralized. Also, its time complexities are approximately O(N^2) in a centralized system and O(N) per peer in a decentralized system.
P2P, free-riding, global contribution, fairness, distributed systems.
H. Nishida and T. Nguyen, "A Global Contribution Approach to Maintain Fairness in P2P Networks," in IEEE Transactions on Parallel & Distributed Systems, vol. 21, no. , pp. 812-826, 2009.