Efficient statistical analysis of microwave circuit performance using design of experiments

Variations of circuit dimensions and material parameters about their nominal values result in microwave circuits that do not perform as expected. The authors describe how the method of design of experiments (DOE) may be used with circuit simulations to statically characterize the effects that variations of these circuit factors have on circuit performance. When applied to circuit simulations, the method allows the design engineer to assess the manufacturability and robustness of a linear microwave circuit design early in the design process. The results of a study that uses DOE for the reflection coefficient performance of an antenna radiator circuit are presented.<>