ABSTRACT
Abstract Performance evaluation of computer and communication networks gives rise to teletraffic models of potentially large numbers of states. Computational analysis of such large-scale models is very demanding in terms of both CPU time and memory requirements. This paper summarizes a unifying and parallelizable system-theoretic approach to efficient solution of a diversity of teletraffic problems, and presents preliminary timing results comparing the sequential and parallel realizations of the approach. These indicate appreciable speed-ups making analysis of large-scale teletraffic problems feasible.
