Immune Agent-Based Intrusion Detection System Model Research

The higher demands are made on the intrusion detection technology with the increasingly noticeable problem of the network security, but the existing intrusion detection has some limitations facing with the fast change of attack technology. Then on the basis of analyzing some representative IDS the shortcomings existing in traditional IDS are presented, such as being lack of distribute distributivity,flexibility,interoperablility.Therefore new technologies are needed for explored to improve the overall performance of IDS.Firstly this paper introduces the research of Intrusion Detection System based on immune principle. Secondly this article especially analyses the usable characteristic of the technology and gives us research revelations, on this basis a new kind of intrusion detection system model is presented based on the immune agent. Based on above discussion a design blue print is provided in detail, which including functions of each module, infrastructure, working flow and algorithms and algorithms of each part and the cooperation scheme among Agent. Expatiated on several aspects such as how to express the network data to antigen, how to produce antibody to detect the antigen, how to organize the Agents to realize distributed intrusion detection and how to response to intrusion, etc.Based on above design, the modules of detection and responding is emulated in the system. Robust and autonomy quality are analyzed toward the model system. Afterwards this thesis evaluate the performance from quantitative and qualitative aspects through experiments. It is proved that this system can effectively raise the ability of measuring and responding in real time and have better robustness and autonomy quality.