P2P Traffic Optimization based on Congestion Distance and DHT

Qian He¹⁺, Qinghe Dong^1,2, Baokang Zhao^{1, 3}, Yong Wang¹, and Baohua Qiang¹

¹Key Lab of Cloud Computing and Complex System,
Guilin University of Electronic Technology, Guilin, 541004 China
{heqian, daphny, wang, qiangbh}@guet.edu.cn

²Guangxi Key Lab of Automatic Detecting Technology and Instruments
Guilin University of Electronic Technology, Guilin, 541004 China

³College of Computer, National University of Defense Technology, Changsha, 410073 China

bkzhao@nudt.edu.cn

Abstract

P2P traffic has brought serious impact on the bearer network, and how to utilize network bandwidth has become a common concern problem for the Internet Service Provider (ISP) and the P2P user. A congestion distance and DHT based P2P traffic optimization method is proposed in this paper. The P2P traffic optimization model imports both aims of the ISP and the P2P user, and the cost of congestion and communication for links is converted into congestion distance uniformly. A distributed scheme with flow controllers and information collectors is realized where information collectors are constructed to a Distributed Hash Table (DHT) overlay. The traffic optimization application case using Bittorrent is discussed, where node selection and choking algorithms are changed. The experiment results show our optimization method can reduce the network traffic of inter autonomous system (AS) substantially, and react to the network congestion automatically, and then the win-win result for the ISP and the P2P user is achieved.

Keywords: P2P Traffic Optimization, Congestion Distance, Bittorent, Win-win

+: Corresponding author: Qian He
Key Lab of Cloud Computing and Complex System, Guilin University of Electronic Technology, Guilin, 541004 China, Tel: +86-773-759-4881

Journal of Internet Services and Information Security (JISIS), 6(2): 53-69, May 2016 [pdf]