Philip J. Kiviat

The statistics of discrete-event simulation

as tracing the elements of a typical experiment from inception through analysis, defining statistical problems encountered and relating them to the formal body of stdtistical theory. The problems described are inherent in all stochastic system simulation models. An experimental design’s ability. to reveal useful insights into a system depends to a great extent on how well these problems are solved. Failure to deal with them may cause errors in interpreting observed associations between system input and output. One common error is the underestimation of the statistical reliability of system response measurements; this error is caused by failure to account for autocorrelation in system response time-series generated by a simulation model. Another frequent source of error is the assumption that random numbers generated within a simulation model are independent, when in fact the method of ran-

The winter simulation conference: perspectives of the founding fathers

In this paper the General Chairs and Program Chairs of the Winter Simulation Conference (WSC) during the period 1967–1974 continue the discussion started in the Keynote Address on the past, present, and future of WSC and the field of simulation.

Keynote Address: 1981 Summer Computer Simulation Conference

Before joining CTEC, Mr. Kiviat was Client Manager of the Washington Office of SEI Information Technology. He was previously Technical Director of the Federal Computer Performance Evaluation and Simulation Center (FEDSIM) and has been associated with Systems Control, Inc., Unicorn Systems Company, Simulation Associates, Inc., The Rand Corporation, and The United States Steel Corporation. He is the

A view from the beginning: when does a description become a taxonomy

This paper describes the authors career leading up to the publication of his 1969 paper digital computer simulation: computer programming languages, how it influenced the paper, and why the paper has endured as a taxonomy for discrete-event simulation programming languages.

Languages for Simulation

Simulation languages are past adolescence and nearing maturity. This is seen in the two papers in this session that describe improved and extended versions of the widely-used simulation languages GPSS and GASP, and by the papers that bring linguistics and mathematics to bear in providing more useful and usable simulation tools.