Reviewed by
Alex Schmid
Savannah Simulations AG, Herrliberg, Switzerland
This book covers a wide range of different topics in computer simulation from basic concepts and tools, continous simulation, discret and combined simulation, fluid dynamics to new simulation techniques like differential inclusion, metric structure in model space or agent-based social simulation. The selection is based primarily, as the author acknowledges, on his own research agenda, and also software examples and case studies are either developed or were conducted by the author. Simulating general relativity or the formation of galaxis, for example, are non-typical case studies that highlight the broad applicability of simulation. The eclectic nature of the book is at the same time its strength and its weakness. Starting from research projects and software developments of the author, many interesting existing and new topics are presented, but as a whole it leaves the reader rather clueless as to what exactly the book adds to existing simulation literature.
The main message of the book seems to be that modeling and simulation is a very generic term for very different activities, techniques and purposes. A cautious remark what regards simulation in general is given by the author with the title "The computer science of illusion". The notion of illusion, even when connected with a truth-seeking intention by the author, seems somewhat ill-chosen from a simulation practitioner to strengthen trust in simulation activities in general.
Apart from these caveats, the book offers an interesting section through what is possible in computer simulation, with respective pros and cons, what regards specific technologies and topics. It also offers some insights and critics of established simulation techniques, for example for discret event simulation (DES). The problem of simultaneous event treating in DES, clearly illustrated by a model of a duel, hints to one specific among other fundamental shortcomings underlying many simulation techniques. In this case, the duel exemplifies the different outcomes of a model, if its „events' are modeled as discret or semi-discret entities.
I will hint to two other more non-typical topics of the book and leave the standards (basic concepts, continous simulation, discrete simulation, optimization) more or less aside. Of specific interest for readers of this journal is the chapter on an agent-based simulation project of terrorist networks (Interactions between hostile hierarchical structures: simulation of the struggle between terrorist and antiterrorist organisations) that was also published in Vol 7 (2) of JASSS. It highlights the specific difficulties of social/agent-based simulation what regards modeling and simulation validation: "to look for a model that simulates real human behavior is utopian". Simulating phenomena of General Relativity is another non-typical example in the book of a simulation application from computational physics, especially chosen by the author to show "illusive" aspects of computer simulation, since some of the relativistic phenomena can be observed and measured, while others are not observable. Simulated topics are light cones, time distortion, approaching a black hole, gravitational lenses and a rotating black hole.
In sum, the book may serve as an eclectic overview of the diversity of topics and tecniques in computer simulation with some interesting applications, a book for people already acquainted with the basic methods and intentions, possibly searching for new ideas for their own simulation practice. It is certainly not an introductory text into computer simulation or a comprehensive overview of computer simulation in general.
Return to Contents
of this issue
© Copyright Journal of Artificial Societies and Social Simulation, 2007