A multi-orthogonal polynomials' approach to bulk queueing theory

Ulises Fidalgo

2023-08-27

Abstract:We consider a stationary Markov process that models certain queues with a bulk service of a fixed number $m$ of admitted customers. We find an integral expression of its transition probability function in terms of certain multi-orthogonal polynomials with respect to a system of distributions that contain measures supported on starlike subsets of the complex plane. We also study the corresponding steady state.

Probability,Complex Variables,Numerical Analysis

What problem does this paper attempt to address?

The paper attempts to address the problem of finding an integral expression for the transition probability function of a stationary Markov process in a batch service queueing theory model where a fixed number \( m \) of customers are admitted for service. Specifically, the author studies this type of queueing model using the method of multiple orthogonal polynomials and obtains an exact integral expression for the transition probability function.

The paper mainly focuses on a batch service queueing model \( M/M(m,m)/1 \), in which customers arrive in batches of fixed size \( m \), and service only begins when the queue size reaches or exceeds \( m \). The goal of the paper is to find the transition probability function for this model and to prove that for \( m > 1 \), all states are transient (i.e., there are no persistent states). Additionally, the paper explores the corresponding steady state and provides the specific form of the transition probability function using the method of multiple orthogonal polynomials.