MODELING STRATEGY AND EVENT PROCESSING LOGIC OF DISCRETE-EVENT SIMULATION LANGUAGES.

Modeling aspects of discrete event simulation using simulation languages have not been studied in systematic manner in the past, despite recent popularity of such simulation languages. This paper evaluates modeling capabilities of two representative simulation languages, GPSS and SLAM, by taking three relatively simple systems, a single server queueing system, a serial 3 machine system, and a polling system. Several distinct but functionally equivalent models are developed for these systems by changing modeling world-view or by changing language capabilities used to model these systems. The developed models are experimentally and conceptually compared, from which important factors to be examined at the time of model development are identified.