论文部分内容阅读
基于NetFlow技术,提出一种基于连接模式的流量检测算法PeerDetector,能够从本地网络的所有流量中有效地检测出P2P流量,PeerDetector算法先用HTD算法检测出HTTP流量,再根据P2P连接的单连接特性识别设计了PTD算法区分开P2P流量和传统网络应用流量,检测出P2P流量.对算法和适合数据处理的数据库进行了设计和实现,用Dynamips仿真Cisco路由器搭建了试验床.经过实验测试,P2P流量检测工具工作效果良好,能够有效检测出绝大多数的P2P流量,误检概率和错检概率非常小.
Based on the NetFlow technology, this paper proposes a PeerDetector based on connection mode, which can detect P2P traffic effectively from all the traffic in the local network. The PeerDetector algorithm first detects the HTTP traffic using the HTD algorithm. Then, based on the single connectivity of P2P connections Identify and design the PTD algorithm to distinguish P2P traffic from traditional network traffic and detect P2P traffic.According to the algorithm and data processing database designed and implemented, Dynamips simulation Cisco router to build a test bed.After the experiment, P2P traffic Detection tools work well, can effectively detect the vast majority of P2P traffic, false positive and false positive detection probability is very small.