Network Traffic Deviation Detection Based on Fractal Dimension

In this paper we examine aggregate network traffic for deviation detection. The precise and fast detection of network traffic deviation is crucial to improve the efficient operation of a network. It is often difficult to detect the time when the defects occur in a network. In this article, a new algorithm is presented to supervise the aggregate network traffic to fast detect the time deviation transpires in a network. This is performed by supervising the statistical attributes of the time series representing the network conduct. The method examines the network conduct using fractal dimension and discrete stationary wavelet transform. In the proposed method, after applying discrete stationary wavelet transform on the signal representing the network traffic, the fractal dimension of the disintegrated signal is computed in a sliding window. Then, variations of signal fractal dimension are considered for deviation detection. Performance of the proposed method is compared with that of three other existing methods using synthetic signal. The results show superiority of the proposed method in terms of accuracy compared to existing methods.