Использование фрактальных свойств трафикав цифровых сетях связи для детектирования сетевых аномалий

Abstract

Article describes problems of network DoS attacks recognition. Detecting method based on evaluation of traffic fractal properties by calculating of Hurst parameter has been proposed. R/S (rescaled range) analysis algorithm was used for its realization. Experimental research of network traffic samples through aforementioned algorithm has been conducted; obtained results reveal that it’s possible to make decision about existence of network traffic abnormalities based on Hurst parameter evaluation.