论文信息 - Experimental Views on Neighbor Selection in BitTorrent

Experimental Views on Neighbor Selection in BitTorrent

BitTorrent is a protocol for P2P content distribution, which has become an important Internet application in recent years. In this paper, we measure the performance of BitTorrent, explain its key elements, and analyze its efficiency. Our contributions are as follows: (1) present an effective way to measure the performance of P2P content distribution systems; (2) quantify Bit-Torrent's Neighbor Selection which impair its efficiency; (3) design a ShareStorm protocol which avoids the BitTorrent flaws. Our preliminary experiment shows that in terms of download time, ShareStorm outperforms BitTorrent by 50+%.

Chanle Wu | Yingchun Lei | Litang Yang | Qi Jiang

[1] B. Cohen,et al. Incentives Build Robustness in Bit-Torrent , 2003 .

[2] Dejan Kostic,et al. Scalability and accuracy in a large-scale network emulator , 2002, CCRV.

[3] Christos Gkantsidis,et al. Network coding for large scale content distribution , 2005, Proceedings IEEE 24th Annual Joint Conference of the IEEE Computer and Communications Societies..

[4] KyoungSoo Park,et al. Scale and Performance in the CoBlitz Large-File Distribution Service , 2006, NSDI.

[5] Pablo Rodriguez,et al. Should internet service providers fear peer-assisted content distribution? , 2005, IMC '05.

[6] John Kubiatowicz,et al. ChunkCast: An Anycast Service for Large Content Distribution , 2006, IPTPS.

[7] Rob Sherwood,et al. Fair File Swarming with FOX , 2006, IPTPS.

[8] William Chan,et al. Improving Traffic Locality in BitTorrent via Biased Neighbor Selection , 2006, 26th IEEE International Conference on Distributed Computing Systems (ICDCS'06).

[9] Rayadurgam Srikant,et al. Modeling and performance analysis of BitTorrent-like peer-to-peer networks , 2004, SIGCOMM 2004.

[10] Venkata N. Padmanabhan,et al. Analyzing and Improving BitTorrent Performance , 2005 .

[11] Ludmila Cherkasova,et al. FastReplica: Efficient Large File Distribution Within Content Delivery Networks , 2003, USENIX Symposium on Internet Technologies and Systems.