Accurate and Effective Nonlinear Behavioral Modeling of a 10-W GaN HEMT Based on LSTM Neural Networks

Abstract

This paper presents a novel nonlinear behavioral modeling methodology based on long-shortterm memory (LSTM) networks for gallium nitride (GaN) high-electron-mobility transistors (HEMTs). There are both theoretical foundations and practical implementations of the modeling procedure provided in this paper. To determine the most appropriate optimizer algorithm for the model presented in this work, four different optimization algorithms are examined. The results of both simulation and experimental validation are provided based on a 10-W GaN HEMT device. According to the developed investigation, the model is capable of extrapolating and interpolating over multiple input power levels and frequencies, including linear, weakly nonlinear, and strongly nonlinear areas. The analysis of the simulated and measured results shows that the developed model has superior performance also when considering the DC drain current (Ids.). Compared with the existing support vector regression (SVR) based model and the Bayesian based model, the proposed approach shows a significantly improved extrapolation capability.INDEX TERMS behavioral modeling, black box model, gallium nitride (GaN), high-electron-mobility transistor (HEMT), long-short term memory (LSTM), microwave frequency, power transistor.

Review of RF Device Behavior Model: Measurement Techniques, Applications, and Challenges

“…Since the 1990s, machine learning has gradually entered the field of RF CAD in the form of ANN [ 48 ]. ANN learns the behavior of RF and microwave devices and circuits through training, combines the knowledge of RF devices with ANN, and uses machine learning methods to improve existing behavioral models and even create new models [ 49 , 50 , 51 , 52 , 53 , 54 ].…”

Automatic multi‐objective particle swarm optimization method for effective Doherty power amplifier design

“…The developed DPA is based on using the GaN HEMT technology, which is well-suited for microwave high-power applications. [30][31][32][33][34] The remainder of the paper is organized as follows. A detailed description of the AMOPSO algorithm is provided in Section 2.…”