A novel estimation of Hurst index based on wavelet

Recent measurement studies show that the burstiness of packet traffic in LAN as well as WAN is associated with self-similar and long-range dependent. In this paper, we analyze them using discrete wavelet transform, and described the nature of the wavelet coefficients and their statistical properties. Then we present an adaptive, efficient unbiased estimation of Hurst index based on multiresolution wavelet analysis and weighted regression. Simulation results based on fractal Gaussian noise (FGN) and real traffic data reveal the proposed adaptive approach shows more accuracy and robustness than traditional methods, which has only O(N) computation. Thus our algorithm can be applied to the real-time application of traffic enforcement and congestion control in high-speed networks.