IMPROVING CALCULATION ACCURACY OF THE SERVICE QUALITY CHARACTERISTICS OF SELFSIMILAR TRAFFIC IN THE NETWORK

Abstract

Researched methods to improve the accuracy of calculation of the characteristics of service quality in the networks with self-similar traffic through more precise location Hurst coefficient depending on the form of Pareto distribution. Since similar traffic (time interval between requests) best describes the Pareto distribution, it is for him gived a new formula for the calculation of the self-similarity coefficient of traffic. The calculation of the characteristics of service quality can be performed on the basis of the Norros formula , which is valid for the model fBM/D/1/∞.