An efficient parameter extraction method using statistical optimization in S-CMOS deep-submicron/nanometer model

An efficient MOSFET modeling methodology with statistical optimization has been introduced to facilitate advanced mixed-signal circuit design. Based on the properties of the S- CMOS model, a combination of local optimization and the group-device extraction strategy is adopted for parameter extraction. By using the statistical method with an Excel solver, a very accurate parameter extraction procedure was developed for the S-CMOS model and was implemented with the optimization software. An accurate parameter set is extracted in accordance with the measurement data of TSMC 0.35-μm technology from MOSIS Service with the LMEE method. The S-CMOS model was implemented into a SPICE3f3 simulator. Several analog and digital circuits were simulated using the popular BSIM3v3 and compared with the S-CMOS model.