Reduced load approximations for loss networks

Abstract

In this paper, we present a general scheme with which to view reduced load approximations in loss networks. We use notation motivated by stochastic Petri net (SPN) representations of such models and a technique similar to that described by Ciardo and Trivedi for general SPNs. Previous reduced load approximations have involved link independence assumptions. In our method, we assume independence between sets of links rather than between themselves. Our independence assumptions are thus less drastic than those that have been made previously and better results can be expected. Several examples are given in this context. © 1993 J.C. Baltzer AG, Science Publishers.