A New Method for Performance Analysis of Overflow Loss Networks

Description

The widely used Erlang's Fixed-Point Approximation (EFPA) has been applied for almost half a century by researchers and practitioners to analyze, manage, design, and dimension telecommunications networks. Despite its wide use, EFPA suffers from inaccuracies that may be significant in certain cases because of simplifying assumptions. Recently, the principal investigator of this project developed a new method called Overflow Priority Classification Approximation (OPCA), which has shown improvements on EFPA for distributed server systems and certain circuit switching applications. The researchers aim to further develop the OPCA methodologies to make it applicable to a wide range of networks, technologies, and topologies. The project will include theoretical and numerical justifications for using OPCA and/or EFPA. This is especially important for Hong Kong where the density of population means that the early adaptation of new network technologies, and the efficient dimensioning of such networks, is important for lower costs for customers and higher financial gains for telecommunication providers.