HEMT statistical modeling using Monte Carlo method combined with Principal Components Analysis

Abstract

A statistical methodology is presented for the extraction of equivalent circuit arameters (ECP's) of high electron mobility transistors (HEMT). This methodology, based upon the Principal Component Analysis (PCA) combined with the Monte Carlo method, can be successfully employed for evaluating HEMT's reliability and their technological dispersion effects, too. Furthermore the method provides an efficient system to derive ECP's values when the measured ones are not available. Thanks to this procedure the device designer can gain a noticeable aid during the design.