Moment estimation of customer loss rates from transactional data

Abstract

Moment estimators are proposed for the arrival and customer loss rates of a many-server queueing system with a Poisson arrival process with customer loss via balking or reneging. These estimators are based on the lengths {S j1] of the initial inter-departure intervals of the busy periods j = 1,..., M observed in a dataset consisting of service starting and finishing times and encompassing both busy and idle periods of the process, and whether those busy periods are of length 1 or > 1. The estimators are compared with maximum likelihood and parametric model-based estimators found previously. ©1998 by North Atlantic Science Publishing Company.