计算机科学 ›› 2014, Vol. 41 ›› Issue (6): 75-78.doi: 10.11896/j.issn.1002-137X.2014.06.015

• 网络与通信 • 上一篇    下一篇

BitTorrent网络节点间流量的数据采集和分析

鄢羽,唐红   

  1. 重庆邮电大学计算机科学与技术学院 重庆400065;重庆邮电大学计算机科学与技术学院 重庆400065
  • 出版日期:2018-11-14 发布日期:2018-11-14
  • 基金资助:
    本文受国家自然科学基金项目(6104004),国家科技重大专项(2012ZX03006001-004),国家973项目(2012CB315806),重庆市自然科学基金项目(CSTC.2009BA2089)资助

Data Collection and Analysis of Traffic Flow between Peers in BitTorrent Networks

YAN Yu and TANG Hong   

  • Online:2018-11-14 Published:2018-11-14

摘要: BitTorrent网络流量占据了互联网总流量的很大比例,吸引了大量专家学者的专注,然而现有实测工具无法实时记录BT网络中节点间的流量交互情况,有关真实BT网络中节点间流量的研究很少。建立了一个BT网络节点间流量的数据采集系统,将其部署在全球性实验平台Planetlab之后得到了真实数据,并验证了数据的正确性。通过对实测数据的分析和研究发现:(1)Tracker服务器在为各下载节点随机返回邻居节点列表时会受到各节点入网时间的影响;(2)50%以上的下载节点从种子处下载的数据为零,大部分下载节点的数据下载对种子的依赖性不大,从而深刻地体现出P2P应用模式的优越性。

关键词: BitTorrent网络,节点间流量,Tracker算法,邻居节点,种子 中图法分类号TP393文献标识码A

Abstract: The fact that BitTorrent networks generate a large portion of the Internet traffic has attracted many experts to focus on it.Unfortunately,due to the existing tools’ inability to measure dynamic volume of traffic between two individual peers,there’s few research on the traffic flow between the peers in real BT networks.The authors designed an inter-peer traffic data collection system,got the real data after implementing it on the Planetlab platform and verified the data’s correctness,analyzed the measured data and found that (1) the return of list of neighbor peers from Tracker server to a downloading peer will be affected by the peers’ joining time into the network,(2) for 50% or more peers,the downloaded data from the seed is zero,and for most peers,the downloading seldom depends on seed,which demonstrates the advantage of the P2P mechanism.

Key words: BitTorrent networks,Traffic flow between peers,Tracker algorithm,Neighbor peers,Seed

[1] Global Internet Phenomena Report:2H 2013 [EB/OL].ht-tps://www.sandvine.com/downloads/general/global-internet-phenomena/2013/2h-2013-global-internet-phenomena-report.pdf
[2] 陈亮,龚俭.大规模网络中BitTorrent流行为分析 [J].东南大学学报:自然科学版,2008,38(3):390-395
[3] Liu Gang,Hu Ming-zeng,Fang Bin-xing,et al.Explaining BitTorrent traffic self-similarity [M].Parallel and Distributed Computing:Applications and Technologies.Berlin Heidelberg:Springer,2005:839-843
[4] 唐红,黄鼎.通用Bittorrent模拟器研究 [J].计算机工程与应用,2011,47(15):130-133
[5] Erman D,Llie D,Popescu A.Bittorrent session characteristics and models [J].River Publishers Series in Information Science and Technology.Special Issue in Traffic Engineering,Performance Evaluation Studies and Tools for Heterogeneous Networks,2009,1:61-84
[6] Erman D,Llei D,Popescu A.BitTorrent traffic characteristics[C]∥Computing in the Global Information Technology,ICCGI’06.International Multi-Conference on.IEEE,2006:42-42
[7] Erman D.Bittorrent traffic measurements and models [M].Blekinge Institute of Technology,2005
[8] 叶明江,吴建平,徐恪.Peer-to-Peer网络流量矩阵的计算模型[J].清华大学学报:自然科学版,2010,50(1):63-66
[9] Xy Ke,Shen Meng,Ye Ming-jiang.A model approach to estimate Peer-to-Peer traffic matrices [C]∥INFOCOM,2011Proceedings IEEE.IEEE,2011:676-684
[10] Tang Hong,Yan Yu,Xing Cong-cong.Traffic modeling and analysis on BitTorrent-like Peer-to-Peer networks [J].Journal of Convergence Information Technology,2013,8(4):173-181
[11] Dale C,Liu Jiang-chuan.A measurement study of piece population in BitTorrent [C]∥Global Telecommunications Confe-rence,2007.GLOBECOM’07.IEEE,2007:405-410
[12] Cohen B.Incentives build robustness in BitTorrent [C]∥Work- shop on Economics of Peer-to-Peer systems.2003,6:68-72
[13] Legout A,Urvoy-Keller G,Michiardi P.Understanding BitTorrent:an experimental perspective(version 3)[M].2005
[14] http://www.planet-lab.org/
[15] http://www.rahul.net/dholmes/ctorrent/
[16] 傅雷扬,王汝传,王海燕,等.R/S方法求解网络流量自相似参数的实现与应用 [J].南京航空航天大学学报,2007,39(3):358-362
[17] 石贵民,林宏基.基于旁路的网络流量监控模式[J].重庆理工大学学报:自然科学版,2011,5(9):63-69

No related articles found!
Viewed
Full text


Abstract

Cited

  Shared   
  Discussed   
No Suggested Reading articles found!