Effective Contract-Net-Based Intrusion Detection Using Intelligent Agent-Based Systems

Intrusion detection is an important task, many times extremely difficult, related to the security policies of an organization that uses digital data and information distributed networks. Intrusion Detection Systems (IDSs) should monitor network traffic and user activity at the same time, to detect malicious traffic and abnormal activity. In this paper, we propose an intelligent agent security approach for adopting IDSs in a distributed computational network-based environment. The solution is a decentralized, agent-based IDS that allocates tasks to knowledge-based agents for collecting, analyzing, and delivering at the same time data needed for identifying threats and for doing effective actions. The advantages of this approach are: scalability, handling increased load and network latency, and no single point of failure.