摘要
研究P2P网络恶意入侵准确检测问题。针对P入侵检测过程与传统的过程不同,采用预兆性特征为主,缺少确切的可确定行为的特征信息,点对点结构限制了特征之间的联系性,传统的特征检测方法主要针对固定特征信息之间的联系性进行入侵判断,一旦遇到点对点特征失联问题,将造成检测不准。提出了一种PSO辨别树的P2P网络入侵检测算法。利用最小二乘法提取P2P网络操作特征,从而为P2P网络入侵检测提供准确的数据基础。构建PSO辨别树,构建点对点特征之间的动态联系,防止特征失联,完成P2P网络的入侵检测。实验结果表明,利用改进算法进行P2P网络入侵检测,能够极大的提高检测的准确性。
Research the accurate detection of P2P network malicious invasion. The paper put forward a P2P net- work intrusion detection algorithm based on PSO identify tree. Least square method was utilized to extract the P2P network operating characteristics, thus providing accurate data for the P2P network intrusion detection. We built PSO to identify trees and built the dynamic characteristic connection between point to point to prevent lost. The we com- pleted P2P network intrusion detection. Experimental results show that the use of the algorithm for P2P network intru- sion detection can greatly improve the accuracy of detection.
出处
《计算机仿真》
CSCD
北大核心
2013年第10期322-324,393,共4页
Computer Simulation
关键词
辨别树
网络入侵检测
网络环境
恶意入侵
Identify tree
Network intrusion detection
Network environment
Malicious invasion
