Network Traffic Anomalies Detection Using a Fixing Method of Multifractal Dimension Jumps in a Real-Time Mode

Abstract

This article considers the possibility of detecting traffic anomalies on the basis of multiscale, multifractal analysis by monitoring fractal-dimensional jumps in real time. The method is based on the current estimation of multifractal properties of traffic using a sliding window and multiscale wavelet analysis. The numerical results allow us to conclude that fixing the jumplike change in the fractal dimension for various components of the multifractal spectrum makes it possible to pinpoint the presence of an anomaly with significant accuracy. In practice, the estimation of these multifractality components in this spectrum can be achieved by constructing a multichannel algorithm, each channel of which is oriented with the corresponding component of the multifractal spectrum.