Abstract
In regard to available features, discrete-event simulation languages differ in many respects. There is, however, one aspect of a simula tion language that transcends all others in significance - namely, the strategy (or strategies) employed for "next event" selection. Strategy determines how a modeler must view a system which is to be modeled (i.e., determines the "world view"). The three generally recognized strategies are event scheduling, activity scan ning, and process interaction. Modern simulation languages may include combinations of these strategies.
Analyses of the three strategies have been made, including the development of algorithms for each of the three strategies, and the use of the algorithms for in-depth comparisons of the char acteristics of languages employing each strategy. Models using different strategies have been compared relative to their develop ment characteristics and performance characteristics. Journal and conference Proceedings' articles spanning many years have been reviewed to obtain observations of simulation experts concern ing the impact of strategy on simulation languages and models.
It is hoped that the perspective of strategies afforded by this discussion can lead to better-informed choices and more effec tive use of modeling tools and methods.
Keywords
Get full access to this article
View all access options for this article.