A novel approach to estimating Hurst parameter for self-similar traffic - art. no. 602216

A variety of traffics from different communications networks have been demonstrated to show the self-similarity. A novel method to estimate the degree of self-similarity based on the index of dispersion for counts, a second-order statistic, is proposed in this paper. Its principle is introduced in brief and validation is made in detail on the basis of two self-similar process models, Sup_FRP and FBNDP. Qualification and comparison show that the method possesses simpler algorithm than the wavelet-based estimator and that it exhibits higher accuracy and reliability than the R/S estimator and the Variance-time analysis.