On Comparison of Multiserver Systems with Two-component Mixture Distributions

IRINA PESHKOVA, Evsey Morozov
15m
In this paper, we introduce the relations between parameters of two-component Hyperexponential, Exponential-Pareto and two-component Lomax mixture distributions which admit stochastic and failure rate comparisons of these mixture distributions. Then we apply the failure rate and stochastic ordering techniques to construct upper and lower bounds for the steady-state performance indexes of a multiserver model with Exponential-Pareto mixture distribution of service times.