Deliberative Agents for Intrusion Detection.

Abstract

This work describes a multiagent system incorporating some artificial intelligence techniques for intrusion detection. The proposed Intrusion Detection System (IDS) provides a network administrator with a comprehensive visualization of the network traffic. Thus, the network manager can supervise the network activity and detect anomalies at a glance. This paper describes the structure of the Mobile Visualization Connectionist Agent-Based IDS (MOVICAB-IDS). The system includes deliberative agents using a connectionist model to identify intrusions in computer networks. Some experiments dealing with anomalous situations related to the Simple Network Management Protocol are described.