LAP Lambert Academic Publishing Distributed Intrusion Detection System: An Evolutionary Computing Approach

This book offers a detailed comparative study on Intelligent Agents-based Distributed Intrusion Detection System (IADIDS) with a fresh research approach. An Intelligent Agent architecture supported by the Evolutionary Computing based technique is proposed for intrusion detection to optimize Distributed Intrusion Detection System (DIDS). The proposed IADIDS architecture consists of two types of Intelligent Agents namely Light Weight Intelligent Agent (LWIA) and Heavy Weight Intelligent Agent (HWIA). Intelligent Agents are distributed across the Distributed Computer Networks. Groups of Intelligent Agents (IA) collectively monitor the network traffic and local activity in order to identify intrusions. IA are organized into sub-environments called LWIA that are connected to each other in a non-hierarchical fashion. Different classes of datasets, knowledge repositories and intelligent classifiers are used to compose the system which performs intrusion detection modeling by employing Intelligent Agents. LWIA implement independent misuse intrusion strategies whose output is systematically fed to HWIA which is capable of more accurate diagnosis and to detect anomaly intrusion also.