Genetic Algorithm Based MOSFET Model Parameter Extraction

The Levenberg-Marquardt (LM) minimization algorithm commonly employed in MOSFET model parameter extraction has several known deficiencies, such as poor convergence characteristics without a good initial guess, low likelihood of convergence to the globally optimal solution, and difficulty with simultaneous multiobjective optimizations. Furthermore, conventional tools require an expert user with a detailed understanding of the MOSFET model and optimization methods to guide the parameter extraction process. In order to overcome these difficulties, an improved genetic algorithm (GA) based minimization technique has been developed. The GA was tested on measured data obtained from nmos devices of five different gate lengths, fabricated using a recent Motorola BiCMOS technology. An advanced surface potential based MOSFET model (SSIM) was used to fit to the data.