Network anomaly behavior detection using an adaptive multiplex detector

Abstract

Due to the diversified threat elements of resources and information in computer network system, the research on a biological immune system is becoming one way for network security. Inspired by adaptive immune system principles of artificial immune system, we proposed an anomaly detection algorithm using a multiplex detector. In this algorithm, the multiplex detector is created by applying negative selection, positive selection and clonal selection to detect anomaly behaviors in network. Also the multiplex detector gives an effective method and dynamic detection. In this paper, the detectors are classified by K-detector, memory detector, B-detector, and T-detector for achieving multi level detection. We apply this algorithm in intrusion detection and, to be sure, it has a good performance. © Springer-Verlag Berlin Heidelberg 2006.