Introduction - Modeling and Simulation of Complex Systems - page 1

1 Introduction
Simulation is nowadays considered to be the third pillar of science, a
peer alongside theory and experimentation [105, p. 12], [55,p.1],[88],
[19, p. 1]. The analysis of many systems, processes and phenomena
is often only feasible by developing simulation models and executing
them using vast amounts of computing power. Forecasts, decision
support and training are further areas that are regularly supported
or even made possible by using simulation.
1.1 Motivation
The basic principle of modeling and simulation is illustrated in Fig-
ure 1.1. The system under investigation needs to be represented as
a model which is suitable for the purpose of the investigation. This
model may then be solved by different means (e. g., analytical meth-
ods or simulation). The results thus gained are finally analyzed and
interpreted with regard to the original system under investigation and
the specific question at hand.
Complex systems are usually characterized by large numbers of
heterogeneous and interacting components resulting in a non-linear
aggregate behavior (i. e., the aggregate behavior is not derivable from
the summations of the activities of individual components) [63]. The
strive for analyzing and understanding ever more complex systems
on one side, and permanently increasing computing power on the
other side leads to more and more complex models. Handling this
increasing model complexity is not really novel, but more of an all-time
challenge. Due to the complexity to be represented within models and
increasingly detailed representation of dynamic behavior, simulation
is often the only choice for analyzing such models.