Adapting and optimization in small signal modeling of nanoscale MOS-FET Using artificial recurrent neural networks

Abstract

In this paper the effect of feedback network structure(EORMLP), in modeling and adapting nonlinear behavior of small signal of nanoscale MOSFETs was investigated,and the optimal accuracy and testing time for determining the device drain current were discussed.we simulate the AC response of BSIM3 modeling of MOSFET with tsmc-0.18um technology , in the frequency of 32Mhz in the inviroment of HSPICE. The training data was generated by using link method from HSPICE to MATLAB 2012.a.finally a good and adapted AINs model were introduced