摘要
针对互联网中常见的VoIP应用类型,重点从分类特征的选取出发,提出了一种基于C4.5决策树算法的VoIP流量识别方法,并分别利用模拟数据和真实网络数据对其进行了测试。结果表明,该方法识别准确率高,对含有不同协议种类VoIP的网络数据具有较强的适应能力。
This paper proposed a novel method to identify several familiar VoIP traffic based on C4.5 decision tree,which was mainly by select the flow features of classification.It tested the simulated data and the true Internet data.The experimental results show the method achieves high indentification accuracy for the Internet data including many different type of VoIP application,and adapt well for the real Internet data
出处
《计算机应用研究》
CSCD
北大核心
2012年第9期3418-3421,共4页
Application Research of Computers
基金
国家“863”计划资助项目(2011AA010605)
关键词
VOIP
C4.5决策树
流特征
VolP（ voice over Internet protocol）
C4.5 decision tree
flow features
