QSIM — A Software Experiment Based on a Microprocessor Evaluation Kit

INTRODUCTION Microprocessors were introduced for the first time into the Department of Electrical Engineering at the University of the Witwatersrand, Johannesburg, in 1976. In that same year one of the authors, M. G. Hartley, on leave from U.M.I.S.T., Manchester, provided a course of lectures on simulation of systems involving stochastic processes. Accordingly, a need arose to provide a laboratory experiment to illustrate the lecture material. The authors, in devising such an experiment, decided to use the opportunity to provide, at the same time, a demonstration of the capability of microprocessors for students new to this subject area. The result is the 'stand-alone' queueing simulation system described in this paper. In discussing the design and implementation of this experiment the authors hope to illustrate how any common microprocessor evaluation kit may be used as the basis of one or other of a series of economical experiments, in addition to its primary function of providing 'hands-on' experience of a small microcomputer system. Physical situations involving a queue or a set of queues are very common. In general, a queue will form whenever the demand for a service exceeds the current capacity to provide that service. The term 'queueing system', therefore, may be applied to a host of different situations where a demand exceeds a supply. Examples of queueing systems include vehicles in a traffic network, people waiting at a post-office counter, warehouse goods awaiting delivery and jobs awaiting processing in a computer system. If such queueing systems were trivial, it would be unnecessary to give them further consideration. In practice, however, the performance of a queueing system is critical. An inadequate service capacity results in long queues which in turn are likely to have economic and social consequences in many practical situations. An excessive service capacity results in unnecessary idle time, which is a direct financial burden, and may also have other far-reaching effects. From this it is clear that, in many