Mobile agent based intrusion detection system adopting Hidden Markov Model

Abstract

Mobile agent based intrusion detection systems distribute detection agents to prevent system from shutting down when an agent becomes breaks. Because in this system agents are distributed, it can reduce network delay and network load, and each agent can operate independently. Agents can also be easily added or deleted. In this paper we propose an enhanced design of mobile agent based intrusion detection system using Hidden Markov Model algorithm for detection. Hidden Markov Model algorithm is used to detect abnormal behavior pattern by analyzing log information. By adopting this algorithm to the mobile intrusion detection agents, detection performance can be improved.