Research On The Intrusion Detection And Protection System Based On Intelligent Agents

With the rapid development of network and computer technology, threat to them are gradually increasing, which are executed in more complex and subtle means, and the security issue has been a matter of great urgency. Compared with firewalls and encryption, intrusion detection is a proactive measure to prevent users from uninvited access, but there are still some flaws in existing Intrusion Detection Systems (IDS): at first, few of them could deal with high-speed network and enormous data easily; secondly, they did not show effective performance on checking new attacks or variational ones in time; another trouble is the high rate of false alert, which result in administrators' over-load. These problems had become chief stumbling blocks to the development and popularization of current IDS.Based on intelligent agents, this paper presents an intrusion detection & protection distributed system model, intending to overcome those disadvantages above. Except most characters of an advanced intrusion detection systems, its distribution, intelligence, real-time response and interactive features are arranged to provide comprehensive network defense and systems protection for users.Because of its proved high speed and efficiency, such system model adopts traditional rule-based misuse detection mode. On this basis, concept learning theory is used to improve abilities of behavior classification and differentiation, rules creation and update, which are to find unknown or variational attacks. For auto generating more effective...