基于机器学习的P2P流量分类系统设计与实现

• 机 构: 湖南工业大学计算机与通信学院;国防科技大学计算机学院
• 出 处: 微计算机信息
• 发表日期: 2009年33期
• 关键词: P2P TRAFFIC CLASSIFICATION MACHINE LEARNING SYSTEM DESIGN AND IMPLEMENT

P2P应用的快速增长,大量消费网络带宽,而传统的基于端口与有效载荷的网络流量分类方法存在着很多缺陷,一种基于机器学习的P2P流量分类系统被建议.系统按照5元组(源IP、源Port、目的IP、目的Prot及IP协议)的定义,将P2P报文分成双向TCP流(Flow)或UDP流,抽取与协议和端口无关的流的特征.用机器学习的方法构造和训练分类器后,对未知协议类型的流进行分类.分类系统的功能模块设计、实现及性能评测被详细的进行了介绍.“,”As the rapid increase of P2P application, a lot of network bandwidth is consumed; but the traditional network traffic classification methods based on port and payload have many objections, so a P2P traffic classification system based on machine learning is proposed. According to the five -tuple definition, the packets are classified into bidirectional TCP or UDP flow. Use machine learning method to construct and training classifier, we can classify unknown application flows. The function design, implement and performance evaluation of the classification system is introduced.