A study of detector generation algorithms based on artificial immune in intrusion detection system

摘要

Detector plays an important role in self and non-self discrimination for intrusion detection system, which makes detector generation a kernel algorithm for artificial immune system. In this paper, firstly current used binary matching rules are listed, characteristics of which are analyzed. And detector generation algorithm is divided into three main processes, including gene library, negative selection and clone selection. Evolution for gene library is explained based on the gene library theory. Several new methods are adopted to improve the performance of NSA, and finally cooperative co-evolution detector generation model is constructed which is a novel structure for intrusion detection system. This paper is aimed for researchers to focus problems on three main ideas concluded in last chapter.