Combined Discrete/Normal Statistical Modeling of Microwave Devices
- 1 October 1989
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
Abstract
This paper deals with statistical characterization of the equivalent circuit parameters for microwave FET devices. The statistics are derived from the S-parameters measured for a sample of devices. The use of a multidimensional normal distribution seems justified for most parameters. For parameters exhibiting sample distributions substantially different from normal we propose a combined discrete/normal approach that can preserve the means, standard deviations, correlations and marginal distributions derived from the sample. This provides enhanced accuracy of the model while retaining the simplicity of the normal distribution. The problem of the size of the sample is also addressed in terms of confidence levels and confidence intervals.Keywords
This publication has 5 references indexed in Scilit:
- An efficient linear statistical FET modelPublished by Institute of Electrical and Electronics Engineers (IEEE) ,2003
- FET model statistics and their effects on design centering and yield prediction for microwave amplifiersPublished by Institute of Electrical and Electronics Engineers (IEEE) ,2003
- Circuit optimization: the state of the artIEEE Transactions on Microwave Theory and Techniques, 1988
- Computer Calculation of Large-Signal GaAs FET Amplifier CharacteristicsIEEE Transactions on Microwave Theory and Techniques, 1985
- Yield optimization for arbitrary statistical distributions: Part II-ImplementationIEEE Transactions on Circuits and Systems, 1980