利用明文特征和朴素贝叶斯分类识别P2P流量

摘要

明文特征是基于应用层静态特征的一种识别方法,需要提取出应用层数据的特征信息;而朴素贝叶斯分类是基于大量统计信息的一种识别方法,主要用来识别加密的Peer-to-Peer(P2P)流量.着重介绍了采用明文特征和朴素贝叶斯分类相结合的方法,对加密的以及未加密的P2P流量进行识别.测试结果表明,这种方法可以较准确地识别出P2P流量.%The mechod of explicit features was based on the static features in application layer, and needed to extract the features of the application layer; the method of naive Bayes classifier was based on the massive statistical information,and was used to identify the encrypted Peer-to-Peer(P2P) traffic.It was the highlights of the paper that the method combined the explicit features and naive Bayes classifier together to identify both of the encrypted and not encrypted P2P traffic.The results show that this method can identify P2P traffic accurately.