Network Intrusion Detection Algorithm Based on Improved Support Vector Machine

Abstract

With the rapid development of Internet and information technology, detecting network intrusion behaviors have been attracted more and more attentions. In this paper, we proposed a novel network intrusion detection algorithm using a hybrid ant colony and support vector machine model. Main ideas of SVM are that it denotes a representation of the examples as points in space, and examples of the separate categories are separated by a gap. Afterwards, framework of the detecting network intrusion system is given, which is designed to promote the accuracy of detecting network intrusion by optimizing parameters of support vector machine with ant colony algorithm. Finally, four types of network attack behaviors are utilized in this experiment, that is, a) DOS, b) R2L, c) U2R, and d) Probe. Experimental results demonstrate that the proposed method is able to detect network intrusion with high accuracy.