A novel artificial neural network‐based nonlinear output current model for GaAs pHEMT with accurate high‐order derivatives

Abstract

In this paper, a novel artificial neural network (ANN)‐based nonlinear output current model with accurate high‐order derivatives is presented. The proposed model can guarantee high precision of the drain current and its high‐order derivatives simultaneously. The existing traditional equation‐based models are analyzed to demonstrate their limits and inaccurate modeling performance of high‐order derivatives. Aiming at maintaining high accuracy of the drain current and its derivatives at the same time, the ANN has been used iteratively. This method, which inherits the advantage of ANN model, obtains higher accuracy than existing ANN model as well as traditional equation‐based models. The effectiveness of the model has been verified by comparing the modeled and measured results for a GaAs pHEMT.

Novel neural network optimization approach for modeling scattering and noise parameters of microwave transistor

“…In the literature, the ANN approach has not only been used in linear behavior modeling for microwave transistors. There are also studies where the ANN approach has been used for DC behavior modeling of transistors 4,20,21 . and large‐signal behavior modeling 22 …”

“…Therefore, transistor modeling studies continue to stay current as an active research field. [1][2][3][4][5][6][7] Scattering (S) parameters are used for linear behavioral modeling of two-port components. 8 The linear behaviors of transistors, which are some of the major components of active microwave circuit design, are determined by S parameters.…”

“…There are also studies where the ANN approach has been used for DC behavior modeling of transistors. 4,20,21 and large-signal behavior modeling. 22 Looking at the literature, it is understood that different ANN architectures have been used for linear behavior modeling of transistors.…”

“…In studies, it is seen that parameters like the numbers of hidden layers, numbers of neurons and activation functions in the structure of the ANNs used have been determined manually with the trial and error method. 4,5,7,10,13,[15][16][17][18][19][20]22 Gunes et al 13 used the General Regression Neural Network (GRNN) and Multi Layer Perceptron Neural Network (MLPNN) methods for modeling transistor S and N parameters. They obtained the MLPNN number of hidden layers, number of neurons in each hidden layer and activation function parameters by the trial and error.…”

A nonlinear empirical I/V model for GaAs and GaN FETs suitable to design power amplifiers

“…Several semi‐empirical analytical I DS ( V GS , V DS ) modeling techniques for GaN and GaAs HEMT devices have been developed, 3‐18 for example, table‐based Artificial Neural Networks (ANNs), or analytical modeling techniques. The table‐based model represents each element to be model with a lookup table developed from measured data.…”