Modeling and Approximation for Statistical Evaluation and Optimization of Microwavedesigns

Abstract

This paper shows how, by suitably updated approximations, one may surmount the obstacle of expensive experimental tuning or repeated computer simulations of trial designs when optimal designs or statistical analyses are required. The authors address themselves to the efficient use of available software by designers wishing to exploit the current state of the art in techniques of statistical design, tolerance assignment and optimal tuning. The ideas and results are new to microwave design.