Abstract: The widespread use of P2P has brought a serious of problems, such as excessive consumption, the rapid spread of the virus and so on. The current research focus is how to identify and control P2P traffic, at the same time people can enjoy the convenience brought by P2P technology. Signature features of the application layer of five mainstream P2P applications are obtained by analyzing TRACE data, and then the algorithm of identifying P2P, based on signature features of the application layer, is proposed in this paper, and the effectiveness of the algorithm is verified by the actual experiment. This research can be directly applied to the identification and management of P2P traffic.

Key words: P2P traffic, application signature, traffic identification, behavioral analysis

摘要：

P2P的广泛使用带来了带宽过量消耗、病毒传播迅速等严重问题。如何在享受P2P技术带来便利的同时又能有效地识别、控制P2P流量是当前的研究热点。通过基于报文TRACE的数据分析方法，获得了当前五种主流的P2P流的应用层签名特征，提出了一个基于应用层签名特征的识别算法，并通过实际实验对所提识别算法的有效性进行了验证，研究成果可直接应用于P2P流量的识别与管理。

关键词: P2P流量, 应用层签名, 流量识别, 行为分析