基于频道签名的P2P-TV流量精细识别  

Fine-grained Identification of P2P-TV Traffic Based on Channel Signatures

作　　者：张鑫[1] 李进[1] 王晖[1] 

机构地区：[1]国防科技大学信息系统与管理学院,长沙410073

出　　处：《计算机工程》2010年第24期9-11,14,共4页Computer Engineering

基　　金：国家"863"计划基金资助项目(2008AA01Z407)

摘　　要：提出一种基于频道签名特征的P2P-TV流量精细识别方法。该方法通过对净流量数据进行统计分析实现频道签名的挖掘,借助频道签名匹配实现精细到单个(IP,Port)、具体到某一平台某一频道的流量识别。针对PPStream、QQLive、UUSee平台的实验结果表明,该方法能有效识别P2P-TV流量对应的平台与频道,且识别精度较高。This paper proposes a fine-grained identification method of P2P-TV traffic based on channel signatures. It mines the channel signatures of popular P2P-TV applications through statistical analysis on their pure traffics. Channel signature matching is used to realize fine-grained identification of the application and channel of the traffic on a particular（IP, Port）. Experimental results on PPStream, QQLive and UUSee prove that the method proposed is effective and accurate.

关 键 词：P2P—TV流量 精细识别 典型平台 频道签名 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]