基于网络流量的P2P动态检测方法  

Dynamic P2P detection method based on network traffic

作　　者：李雯[1] 梁刚[1] 

机构地区：[1]四川大学计算机学院,成都610065

出　　处：《四川大学学报（自然科学版）》2011年第2期315-320,共6页Journal of Sichuan University(Natural Science Edition)

基　　金：国家自然科学基金(60873246);教育部创新工程重大项目培育资金(708075);国家教育部博士点基金(20070610032)

摘　　要：随着目前的P2P应用开始采用随机端口、隧道和加密等技术,传统检测方法的识别率逐渐降低.针对这一问题,提出了一种基于网络流量的P2P动态检测方法.该方法通过分析网络流量对不同检测方法的误报率和漏报率的影响,采用流量阈值机制动态地选择检测方法,并通过将负载均衡的思想运用于流量阈值的调整,实现了一种变步长的自适应算法,提高了适应性和检测效果.实验结果证明,该方法可以稳定地、高效地识别不同网络环境下的P2P流量.With the development of P2P applications, some technology has emerged, such as random ports, tunneling, encryption and so on, which decreased the identification rate of traditional detection methods. A dynamic P2P detection method based on network traffic was proposed to this problem, and traffic threshold which selected detection algorithms dynamically was adopted to improve adaptability and detection performance, with analyzing the influence of network traffic on false positive rate and false negative rate of different detection algorithms. What＇s more, load balancing was applied to adjustment of traffic threshold to realize a variable step-size adaptive algorithm, and the experiments proved that it made P2P identification in diverse networks stable and efficient.

关 键 词：P2P流量检测 流量阈值 负载均衡 动态 自适应算法 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]