Intrusion Detection System Based On Distributed Intelligent Agents

With the popularization and fast development of Internet , network users face increasingly serious security issues, thus network intrusion has became the most important threat to the computer security and network securityo Intrusion detection as a new fast growing field has become one of the most important research directions in the network security research fields。Intelligent agent technology is a new kind of software designing mode。 Intelligent agents are autonomous,interactive and mobilable, Thus agent technology provides a novel and feasible design scheme for network based distributed applications include intrusion detection。 So it is very worthy and significant in both academe and practive to do research work in the intelligent agent based on distributed intrusion detection field oIn this dissertation, we investigate problems in the distributed intrusion detection systems based on intelligent agent technology , and give a novel kind of IDS-ADIDS。The main work and novel parts of this dissertation are:Firstly, overcoming the problem of single point invalidation and the bottleneck of treatment ability in hierarchical system, we uses a method of compound components。 For example, the tasks of an intelligent agent can be done completely by the other intelligent agents in ADIDS。Secondly , each kind of intelligent agents only served as some special tasks of the intrusion detection , and intelligent agents needn't owe to the all functions of intrusion detection o .So wo assemble these intelligent agents with the functional components from functional components pools to meet the needs of the intrusion detection。 This method not only can reduce the resources consumption of the key computers and network., but also can raise efficiency of the intrusion detection system-ADIDS。Thirdly , ADIDS makes full use of Snort's plug-in components in constructing functional components pools of ADIDS , thus these intelligent Agents conveniently...