Measurement and modeling of the peer-to-peer sharing objects popularity

GUO Zhen-bin; QIU Zheng-ding

您当前的位置：

首页 >

文章列表页 >

Measurement and modeling of the peer-to-peer sharing objects popularity

更新时间：2024-10-14

- Measurement and modeling of the peer-to-peer sharing objects popularity
- Vol. 30, Issue 3, Pages: 66-74(2009)
- 作者机构：
  
  北京交通大学信息科学研究所
- 作者简介：
- 基金信息：
- DOI：
  CLC： TP393.093
- Published：2009
- 稿件说明：
移动端阅览
GUO Zhen-bin, QIU Zheng-ding. Measurement and modeling of the peer-to-peer sharing objects popularity[J]. 2009, 30(3): 66-74.
DOI：

GUO Zhen-bin, QIU Zheng-ding. Measurement and modeling of the peer-to-peer sharing objects popularity[J]. 2009, 30(3): 66-74. DOI：

摘要

使用主动探测BitTorrent tracker服务器的方法

对4个BitTorrent tracker服务器的资源共享流行度和资源下载流行度进行了测量与建模

并使用χ2拟合检验法对模型进行了检验。发现资源流行度具有严重的不均衡性:最流行的20%资源具有约60%的共享

而最流行的10%资源具有60%以上的下载。进一步解决了目前P2P资源流行度研究中存在的Zipf模型是否适用的争论

得出以下结论:①Zipf模型和Zipf-Mandelbrot模型都适合于对资源共享流行度进行建模

而Zipf-Mandelbrot模型比Zipf模型建模的效果更好;②Zipf模型不适合于对资源下载流行度进行建模

而Zipf-Mandelbrot模型具有完美的建模效果。

Abstract

A detailed study on the BitTorrent sharing objects popularity was presented.The objects-sharing popularity and objects-downloading popularity of four BitTorrent trackers was measured and modeled by means of actively probing the tracker server

and the model was tested by χ2 fitting test.The experiment results indicate that the peer-to-peer sharing objects popularity is very unbalanced

that is

there are almost 60% of total share copies for the most 20% popular objects and there are above 60% of total downloads for the most 10% popular objects.Moreover

the experiment results also indicate that the objects-sharing popularity can be well modeled by Zipf distribution and perfectly modeled by Zipf-Mandelbrot distribution

and that the objects-downloading popularity can not be modeled by Zipf distribution and can be perfectly modeled by Zipf-Mandelbrot distribution.

关键词

Keywords

references

Views

931

下载量

CSCD

Alert me when the article has been cited

提交

Tools

Publicity Resources

Research on the IP alias resolution technology

Risk assessment model based on fuzzy Petri nets

Smart-blacklisting:an efficient methodology for mitigating fake block attack in P2P file-sharing systems

MP-Start:bandwidth measurement based and phase-divided TCP slow start mechanism

Evaluating the network security by the Read_Write transition model

Related Author

Huaping CAO

Nianfeng WENG

Han LI

Chao HU

Zhanfeng WANG

Guang CHENG

Xiang GAO

Yue-fei ZHU

Related Institution

School of Computer Science and Engineering, Southeast University

College of Cyberspace Security,Southeast University

College of Comannd Control Engineering,Army Engineering University of PLA

National Computer Network Emergency Response Technique Team,Coordination Center of China

Nanjing Telecommunication Technology Research Institute

AI问答

⁰