A NOVEL METHOD OF INTRUSION DETECTION BASED ON ARTIFICIAL IMMUNE SYSTEM
This paper presents a novel method of intrusion detection based on artificial immune system. Adopting the constraint-based detectors and any-r intervals matching rule, a novel solution is presented to encode the antibody-antigen. Some immune related concepts are introduced. In order to accelerate the accessing of the normal IP packets, the self-pattern class is proposed. Using the data sets of KDD CUP1999, the experiment results show that the proposed method can achieve a faster running speed and better detecting rates. It also can adapt to dynamically changing environments.
Antigen-antibody constraint-based detectors any-r intervals matching rule self-pattern class
YU-FANG ZHANG GUI-HUA SUN ZHONG-YANG XIONG
College of Computer, Chongqing University, Chongqing 400044, China
国际会议
2006 International Conference on Machine Learning and Cybernetics(IEEE第五届机器学习与控制论坛)
大连
英文
1602-1608
2006-08-13（万方平台首次上网日期，不代表论文的发表时间）