Simulation based investigation of conservation laws in queuing networks

Abstract

A queuing discipline is nothing more than a means for choosing which customer in the queue is to be serviced next. The selection may be based on policies like first come first serve (FCFS), shortest job first (SJF), head of the line (HOL) etc. The last one is usually referred to as a priority queuing discipline whose effect has been studied based on some specific conservation laws using the simulation models for a semiconductor fab. The developed model has been simulated on a simulation package called AutoSched AP. The performance of the simulation model has been analyzed by changing various parameters including control policy, product mix, no. of servers, loading on bottleneck toolsets, and probability distribution. In addition to this, two different conservation laws were studied in the context of priority queuing networks. It was observed that little’s law corollary which is independent from the no. of servers and their respective loading performed better in estimating the cycle time changes in different product classes of priority queuing network.