By using grey system and Neural-Fuzzy Network methods to obtain the threshold voltage of submicron n-MOSFET DUTs

In this paper, two techniques are used to obtain the complex non-linear threshold voltage (Vth) data in sub-micrometer MOSFET DUTs via the grey system (GS) method and Neural-Fuzzy Network (NFN). This paper presents the implement procedure of these two models in Vth predictions. Moreover, comparisons among the GS and NFN output results are carried out. Here, it will be used to analyze the Vth inclination of submicron MOSFET DUTs due to the device geometric effect. Introducing comparison between the measured and prediction characteristics of Vth show good matching for a wide domain of channel width (W), length (L), and bias conditions. Then, the implemented procedure may be proper for BSIM model parameters extraction.