Performance Analysis Of A Wavelet-Based Hurst Parameter Estimator For Self-Similar Traffic

In this paper we evaluate the performance of a wavelet-based estimator for the estimation of the Hurst parameter used in characterizing self-similar network traac. We use genuine traac traces and simulation tools to investigate the impact of short-range dependence on the accuracy of the estimated parameter. We brieey address the implication of our ndings on traac engineering. Our main conclusion is that this wavelet-based estimator can be used to analyze aggregated traac data. Nevertheless, the es-timator proves to be unsuitable for the medium-and high-bursty video traac because of the traac's complex correlation structure.