Network anomaly detection based on signal processing techniques

Abstract

Abstract The article depicts possibility of using Matching Pursuit decomposition in order to recognize unspecified hazards in network traffic. Furthermore, the work aims to present feasible enhancements to the anomaly detection method, as well as their efficiency on the basis of a wide collection of pattern test traces.