Simulating computer systems

Simulation is a good way to investigate performance and behavior of computer systems/networks under various input conditions. The importance for simulation in studying such systems includes: impracticality of dealing directly with the system/network; cost of studying the system/network directly, which may be too high; and the complexity of a mathematical model precluding any possibility of an analytical solution. A reasonable approach to obtaining a manageable model of the system consists of abstracting its features and properties. Particular attention is paid to those pertinent to the study. When we are doing modeling and simulation, we are concerned not only with the elements viz. the real system, model and computer, but also with establishing certain relationships among them. In particular, modeling deals primarily with the relationships between real systems and models. This article is intended to provide a broad overview of various aspects involved in simulating computer systems and assumes a basic knowledge of probability distributions. The simulation of a single server queueing system is also considered