摘要
本文针对BitTorrent共享文件类P2P应用,提出了一种混合式的流量检测方法。该方法由三个子方法构成,分别针对BitTorrent流量中的明文流、密文流和信令流进行检测,并预知即将发生的BitTorrent流量。实验结果表明,该方法的召回率、准确率和实时性,均优于目前实时性最好的几种机器学习方法。
in this paper, according to the BitTorrent file sharing P2Papplication, presents a flow detectionmethod of mixed type. The methodconsists of three sub methods, respectively, for BitTorrent flowin plaintext stream, stream cipher and signaling flow were detected, and predict theBitTorrent traffic coming. The experimental results show that, the rnethod of the recall rate, accuracy and real-time, are superior to the best real timeseveral machine learning methods.
出处
《网络安全技术与应用》
2014年第3期53-55,共3页
Network Security Technology & Application
关键词
流量检测
应用层签名
消息重组
信令分析
traffic detection
application layer signature
messagerecombination
signal analysis
