Selection of Plan of Experiment by Statistical Analysis of the Parameters of Teletraffic Model with Voice Services

Abstract

A teletraffic model of voice services had been created. Simulations of the model of composite symmetric Bm, symmetric quasi -D optimal and non-symmetric quasi -D optimal types of experimental plans have been carried out. The plans of the experiment and the received experimental data are presented. Check on the suitability of the regression model of second degree has been done. The results of applied regression analysis are presented. The best plan of the experiment has been selected. The most suitable mathematical models on the average downtime of the system and the probability of losses have been found.

Estimation and Optimization of Worm Drive Efficiency using Taguchi and Symmetric Quasi-D Optimal Methods

“…It is often more convenient and economical to plan and implement a large series of experiments at once than several small series in succession, particularly in cases where the response surface in the factor space being studied is nonlinear. In many cases, the use of a noncompositional plan is associated with higher efficiency, such as when the response surface area is known to have nonlinear characteristics [24,25]. The Box-Behnken design and Pesochinsky's symmetric quasi-D optimal methods of experiments can be used to model the response surface and is not based on FFE or FRFE.…”