Improving of the security of intrusion detection system

Abstract

Agent-oriented hardware and software are proposed to use in order to improve security of intrusion detection system. A generalized structure of intelligent intrusion detection system was developed. A neural network was used as its main agent. Neural network immune detectors were implemented on the programmable logic arrays in order to detect and classify intrusions. Mamdani fuzzy inference rules were applied to counteract intrusions. A decision-making subsystem structure was developed.