CAD techniques for robust RF and wireless IC design

Statistical CAD methodology for wireless RF IC's is presented. Orthogonal-Array design of experiments screens and quantifies the effects of parameter variations on circuit performance. Latin Hypercube Sampling, Maximally Flat Quadratic Interpolation (MFQI), and Group Method of Data Handling (GMDH) algorithms create performance approximation models. Models are used in optimization which optimize and minimize design variability. 2.4 GHz LNA is utilized to demonstrate the viability of the methodology.