APPROXIMATION OF PROBABILITY DISTRIBUTIONS IN QUEUEING MODELS

T. I. Aliev


Read the full article 
Article in Russian


Abstract

For probability distributions with variation coefficient, not equal to unity, mathematical dependences for approximating distributions on the basis of first two moments are derived by making use of multi exponential distributions. It is proposed to approximate distributions with coefficient of variation less than unity by using hypoexponential distribution, which makes it possible to generate random variables with coefficient of variation, taking any value in a range (0; 1), as opposed to Erlang distribution, having only discrete values of coefficient of variation.


Keywords: probability distribution, coefficient of variation, approximation, hypoexponential distribution, hyperexponential distribution, Erlang distribution
Copyright 2001-2017 ©
Scientific and Technical Journal
of Information Technologies, Mechanics and Optics.
All rights reserved.

Яндекс.Метрика