TABLE OF CONTENTS Many real-life systems exhibit behavior with random elements. Such systems encompass a vast array of application areas, such as the following:
Manufacturing
Random demand for product held in an inventory system
Random product processing time or transfer time
Random machine failures and repairs
Transportation
Random congestion on a highway
Random weather patterns
Random travel times between pairs of origination and destination points
Telecommunications
Random traffic arriving at a telecommunications network
Random transmission time (depending on available resources, such as buffer space and CPU)
Indeed, simulation modeling with random elements is often referred to as Monte Carlo simulation, presumably after its namesake casino at Monte Carlo on the Mediterranean. This apt term commemorates the link between randomness and gambling, going back to the French scientist Blaise Pascal in the 17th century.
Formally, modeling a random system as a discrete-event simulation simply means that randomness is introduced into events in two basic ways:
Event occurrence times may be random.
Event state transitions may be random.
For instance, random interarrival times at a manufacturing station exemplify the first case, while random destinations of product units emerging from an inspection station (possibly needing re-work with some probability) exemplify the second. Either way, probability and statistics are fundamental to simulation models and to understanding the underlying random phenomena in a real-life system under study. In particular, they play a key role in simulation-related input analysis and output analysis. Recall that input analysis models random components by fitting a probabilistic model to empirical data...
