Modified Algorithm for Detecting Network Attacks Using the Fractal Dimension Jump Estimation Method in Online Mode

The paper considers a modification of the well-known algorithm for detecting anomalies in network traffic using a real-time fractal dimension jump estimation method. The modification uses real-time thresholding to provide additional filtering of the estimated fractal network traffic dimension. The accuracy of the current estimate of the fractal dimension and the reliability of anomaly detection in network traffic in online mode is improved by adding extra filtering to the algorithm.