Exploring Artificial Neural Network for X-Parameter and S-Parameter Modelling of HEMT

In this research paper, we have explored the efficiency of artificial neural network (ANN) for the S-parameter and X-parameter modelling of high electron mobility transistors (HEMTs), which are crucial elements in the linear and nonlinear characterization of HEMT devices. We used a GaN HEMT device to obtain X-parameters and S-parameters. The obtained data is used to train an ANN-based model that is validated for interpolation and extrapolation. Experimental results indicate that the ANN-based model accurately predicts the S-parameters, showcasing its efficacy in characterizing the device's linear behaviour. However, the performance of the ANN model in predicting the X-parameters is found to be less efficient and the result is supported by data analysis and graphical representations, highlighting the need for additional research to improve the ability of the ANN model to accurately predict the nonlinear behaviour of HEMTs.