Discrete-Event Simulation: A First Course

Simulation is a popular and valued analytical technique. Quite often, surveys of simulation practitioners list simulation among the top techniques in popularity and in use. One’s perspective of simulation may be viewed along a continuum: from the perspective of a computer sciencefocused developer of simulation code to the perspective of a management science-focused user of simulation output (the computer science, management science qualifiers are of course notional). Supporting this continuum of simulation perspective, a simulation course may focus on how to build simulations from first principles, such as from the computer science perspective or may focus on how to use simulation output to provide insight for decision making, such as from a management science perspective. The Lawrence M. Leemis and Stephen K. Park (2006) text, ‘Discrete-Event Simulation: A First Course’ is an excellent simulation text targeting the detailed modelling, computer science-focused perspective of simulation modelling and course focus. At a minimum it should reside on the reference shelf of any simulation professional. As a course text, it warrants serious consideration for any course whose objectives are focused on the details of simulation modelling. The text is well suited to any simulation course focused on the ‘nuts and bolts’ of simulation modelling. As noted, the authors encourage ‘self-discovery’ of key simulation modelling concepts through the use of a general-purpose language and author-provided programs (in C and Java). The text focus is on simple simulation systems. This contrasts with the model-building-focused texts such as Arena (Kelton et al, 2007), with its emphasis on building more realistic systems from a library of black-box components such as found in commercial simulation packages. The natural question is how this text compares to such standard texts as Law and Kelton (2000), now just Law (2007), or Banks et al (2005). In short, this Leemis and Park text is more focused and less comprehensive than the standard texts (and they do refer to these standard texts for further details). However, the limitations seem quite justified when viewed in the context of creating a details-focused course on discrete-event simulation. It was therefore not a distracting feature of the text. Chapter 1 is an excellent introduction to the simulation modelling process. While every simulation text contains such material, this version flows better than most, particularly with the authors’ algorithmic depiction of the simulation process, and provides a nice conceptualization of the process. Chapter 2 addresses random number generators and their underlying theory, providing more depth into the subject than found in most simulation texts, including the more comprehensive broad coverage found in Law (2007) or Banks et al (2005). The level of details provided is often found only in specialty texts and, as such is a definite read/ review for simulation professionals and anyone preparing to teach a class on simulation. Chapter 3 is called discrete-event simulation, but really focuses on generating random number streams as a means to drive a computer simulation model (as opposed to tracedriven simulation). In keeping with their small-model focus, this chapter provides insight into the mechanisms within the simulation process. Every simulation text seems to include a statistics refresher chapter (sometimes as an appendix). In Chapter 4, this refresher is focused on the statistics needed to conduct simulation output analysis, both from a quantitative and a qualitative aspect. The chapter does appear to give particular emphasis to the important problem of simulation output data correlations. Chapters 5–7 give a fairly detailed coverage of how to implement next-event scheduling (ie list management) followed by discrete and then continuous random variable Journal of Simulation (2007) 1, 147–148 r 2007 Operational Research Society Ltd. All rights reserved. 1747-7778/07 $30.00