A novel intrinsic-parameters-correlation enhancement technology applied to accurately extract GaN HEMT small-signal model parameters

Abstract

In this paper, a novel intrinsic-parameters-correlation-enhancement method combining Principal Component Analysis (PCA) algorithm and Particle Swarm Optimization (PSO) algorithm for extracting intrinsic parameters of GaN High Electron Mobility Transistors (HEMT) is proposed. The traditional intrinsic parameter extraction methods are time-consuming and have low accuracy for modeling S-parameter. In order to improve the model accuracy, the PSO algorithm can be used to optimize the intrinsic parameters. However, the PSO algorithm does not consider the correlation of the intrinsic parameters, which leads to a limited improvement in model accuracy. To further improve the model accuracy, in this paper, the PCA algorithm is used to process the real and imaginary parts of the intrinsic model Y-parameters, which can enhance the correlation between intrinsic parameters. Then, the new intrinsic model Y-parameters and the PSO algorithm are used to extract the intrinsic parameters. To validate the effective of the proposed technology, it is applied to extract GaN HEMT small-signal model parameters in the frequency range of 0.5–20.5 GHz, and the experimental results show that the S-parameter modeling accuracy is effectively improved.