Simulation model for extended double-ended queueing

The purpose of this paper is to extend traditional double-ended queuing models using a simulation approach. Traditional double-ended queuing models assume that one supply queue should satisfy one demand queue through instantaneous pairing. Inter-arrival time is assumed to follow an exponential distribution, with arrivals to the system assumed to occur just one at a time. However, this assumption is frequently violated in many real-world situations. The pairing or batch size can either be multiple or a random variable, and the pairing processing time can be greater than 0. Inter-arrival time may follow distributions other than exponential. In some cases bulk arrivals may come at the same time, and pairing is not always guaranteed. Because the analytical approach has enormous difficulties obtaining performance measures under these relaxed situations, a simulation approach for extended double-ended queueing processes is presented. This includes an algorithm to find state probabilities and a newly developed simulation procedure. Using this new procedure, sensitivity analyses of performance measures were performed using various input conditions implemented using ProModel and SimRumnner simulation software. A business case is studied to demonstrate the versatility of the proposed approaches.