A multi-agent system for computer network security monitoring

Due to the complexity of the problem of the network security diagnosis it is necessary to apply effective mechanisms allowing coping with not only knowledge enormity, but also incompleteness and inconsistency. In this connection a proper solution is to apply the multi-agent approach in the network security systems. In this paper the multi-agent system for computer network security monitoring is presented. The overall architecture including agents' knowledge representation, knowledge management and behavior specification is proposed.