Variation of Access Control Policy for Two-queue Model

Abstract

We consider access control problem for two multi-server loss queues in tandem, which is usually used to model networking systems. In previous work [4], it can be shown that under appropriate conditions the optimal managerial policy that maximizes the expected total discounted reward over an infinite horizon is given by a switching curve in the two-dimensional state space. In this paper, we discuss the variation of these switching curves with number of customers in system. Some experimental examples of unusual variation are presented. However, two sufficient conditions, which prevent this counterintuitive situation from happening were proposed.