Wavelet-Based Real Time Detection of Network Traffic Anomalies
Real time network monitoring for intrusions is offered by various host and network based intrusion detection systems. These systems largely use signature or pattern matching techniques at the core and thus are ineffective in detecting unknown anomalous activities. In this paper, the authors apply signal processing techniques in intrusion detection systems, and develop and implement a framework, called Waveman, for real time wavelet-based analysis of network traffic anomalies. Then, they use two metrics, namely percentage deviation and entropy, to evaluate the performance of various wavelet functions on detecting different types of anomalies like Denial of Service (DoS) attacks and portscans. Their evaluation results show that Coiflet and Paul wavelets perform better than other wavelets in detecting most anomalies considered in this paper.