Browse Books

Reviewer: David H. Withers

This could have been a great book, but there are editorial problems that reduce it to something between good and very good. The great part is the technical topic and the formalisms for modeling that are the main content of the text. The methods presented in this book can be used by serious researchers in modeling and simulation to make a substantive contribution, especially in discrete event simulation and modeling. The book is divided into five parts: (1) an introduction and motivation, (2) the fundamentals for the formalisms that appear later, (3) modularity and hierarchy for models (Ziegler calls this area multicomponent models), (4) system modeling with experimental frames and model construction, and (5) architectures for mode- ling. The author claims the book can be used for a graduate-level course in com- puter or systems science, using a selected subset of the chapters. I doubt that students without experience in modeling of nontrivial systems will derive the intended benefits. I feel that researchers in discrete event simulation are in the best position to take advantage of the content of the book. Some very astute students will be able to use the material, but only with considerable effort. I have a few concerns about the technical content. They do not substantially detract from the text. Ziegler assumes that a good decomposition creates “plugs” and “sockets.” This is probably true of hierarchical decompositions, but I'm left wondering if it is a result in general. He does not include the concept of migrating automatically from a simulation to an operating system. He claims that the time base is required to be a linearly ordered Abelian group under addition. This may be true, but it's not obvious. Table 3.1 is meant to be some examples of various formulations (differential equations, discrete event, etc.), but the discussion could lead a reader to the conclusion that the list is exhaustive rather than just examples. Chapter 14 is not as effective as it could be. There is too much abstraction without supporting examples for the students. The key results in the text are found in Chapters 8, 9, 11-13, 15, and 16. In Chapter 8, he shows how the coupling of Discrete EVent Simulation (DEVS) component models results in a composite system that is also a DEVS. Chapter 9 presents four ways to simplify models and the conditions under which valid models will result from simplification. Chapter 11 presents the concept of an entity structure. In Chapter 12, the need to separate the model and the experimental frame is made clear. Chapter 13 extends experimental frames and defines an ordering principle for them. Chapter 15 develops a methodology for model building that may lead to automation of the process. Chapter 16 shows a prototype programming methodology for model building. There are some minor editorial errors that detract from the text: “spaghettization” is used on p. 10; “concretization” is attempted on p. 22; the use of time and frequency analysis is reversed on p. 14; Figures 1.5 and 1.6 don't match the text in typestyle and use of upper/lower case; and a paragraph on p. 22 begins with “thus,” but there's nothing in the paragraph that concludes from prior information. Also on p. 22, we are asked to consider the following example. However, three examples and one figure follow, and the reader is not sure what we are meant to consider. The formatting in Chapter 2 is generally very confusing (pp. 25–27 are particularly bad). The confusing discussion about finite state machines should be replaced with a more cogent discussion of an example that would be familiar to simulation modelers. New notation is introduced on p. 30 without definition. Some chapters have no exercises, some have exercises scattered randomly throughout the text, and Chapter 13 has a good set of exercises at the end of the chapter. Solutions to the exercises are not included. Figure 3.1 is wasted because a whole world of new notation is introduced without explanation. Also, p. 44 is cluttered with new symbols that are used without definition. Chapter 4 is an obvious replica of a paper written for a journal. The style is very stiff, very few examples are included, and there is no motivation for the student. There is a complete list of prior work at the beginning of the chapter. Typesetting inconsistencies on pp. 98 and 99 tend to confuse the reader. Brackets are need for the equations on p. 164, and “demon” is supposed to describe something on p. 215. The editorial and formatting problems could be easily corrected. I would encourage the author and the publisher to consider a second edition soon. The content is too valuable to be obscured by small things that confuse the readers—especially student readers.