Application of Artificial Neural Networks in Noise Wave Modeling of DG MESFETs

Abstract

An improved procedure for noise modeling of dual- gate MESFET (DG MESFET) based on transistor noise wave model is proposed in this paper. A DG MESFET small-signal model is composed of two cascoded single-gate MESFET intrinsic equivalent circuits embedded in a network representing the device parasitics. The wave interpretation of the device noise is used for defining the noise parameters of each single gate MESFET. An artificial neural network is included in the noise model in order to model the noise parameters more accurately.