A Synthetic Dimension Reduction in Intrusion Detection System

Abstract

In order to improve the performance of Intrusion Detection System (IDS), a synthetic dimension reduction method is proposed in this paper. First of all, we define a similarity distance algorithm between two vectors based on analogy reasoning. Then, the merit of the synthetic dimension reduction is analyzed in a 3-dimension space. Finally, the distances between a new behavior sample which is sniffered from network and behavior sample sets. Finally, using these two distances as ordinate and abscissa, this new behavior sample is mapped into a point in a two-dimensional coordinates plane from a multi-dimensional vector space. According to the location of this point, an behavior can be determined whether it is a intrusion.