Optimized Machine Learning-Augmented Hybrid Empirical Models for AlGaN/GaN HEMTs: A Comprehensive Analysis
Precise modeling of gallium nitride (GaN) high-electron mobility transistors (HEMTs) is vital for ensuring reliable and scalable RF circuit design, and efficient characterization of the device behavior. This article presents robust hybrid equivalent circuit (EC)–machine learning (ML) fram...
Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2025-01-01
|
| Series: | IEEE Access |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/11105079/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1850034099854508032 |
|---|---|
| author | Ahmad Khusro Saddam Husain Mohammad Hashmi |
| author_facet | Ahmad Khusro Saddam Husain Mohammad Hashmi |
| author_sort | Ahmad Khusro |
| collection | DOAJ |
| description | Precise modeling of gallium nitride (GaN) high-electron mobility transistors (HEMTs) is vital for ensuring reliable and scalable RF circuit design, and efficient characterization of the device behavior. This article presents robust hybrid equivalent circuit (EC)–machine learning (ML) frameworks for significantly streamlining the extraction of small-signal model parameters of AlGaN/GaN HEMTs. The extrinsic and intrinsic parameters of the devices are initially extracted using physics-relevant empirical models in Keysight’s advanced design system. Thereafter, six extensively optimized ML regression models, namely decision tree (DT), ensemble learning (EL), support vector regression (SVR), kernel approximation regression (KAR), Gaussian process regression (GPR), and neural networks (NN) are employed to simulate the intrinsic behavior of GaN HEMTs. The models are trained on GaN HEMTs of geometries <inline-formula> <tex-math notation="LaTeX">$4\times 100~\mu $ </tex-math></inline-formula>m, <inline-formula> <tex-math notation="LaTeX">$10\times 220~\mu $ </tex-math></inline-formula>m, and <inline-formula> <tex-math notation="LaTeX">$10\times 250~\mu $ </tex-math></inline-formula>m, while tested on GaN HEMTs of geometries <inline-formula> <tex-math notation="LaTeX">$2\times 200~\mu $ </tex-math></inline-formula>m and <inline-formula> <tex-math notation="LaTeX">$10\times 200~\mu $ </tex-math></inline-formula>m across diverse biasing and frequency conditions. The input features to models include gate–source voltage (VGS), drain–source voltage (VDS), frequency, number of fingers (NF), unity gate width (Wg), and effective gate width (Weff). Finally, a thorough quantitative assessment and detailed comparisons are performed in terms of standard regression tests, mean absolute percentage error, Nash–Sutcliffe efficiency, Kling–Gupta efficiency, training and prediction speed, reliability of model parameters, and simulation agreement with the measured S-parameters. The results demonstrate that among the tested ML models, EL exhibited the lowest mean relative S-parameter errors (2.78–3.75 %), followed by NNs (2.05–6.98 %), DT (2.29–7.35 %), GPR (2.86-8.91 %) SVR (7.83–9.88 %), and KAR (8.26–10.54 %) across diverse GaN HEMTs geometries. This hybrid modeling strategy provides a practical alternative to conventional parameter extraction, offering speed, accuracy, and broader applicability. |
| format | Article |
| id | doaj-art-2c1fd41e421a4cf28f8c308da39e8a1c |
| institution | DOAJ |
| issn | 2169-3536 |
| language | English |
| publishDate | 2025-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Access |
| spelling | doaj-art-2c1fd41e421a4cf28f8c308da39e8a1c2025-08-20T02:57:57ZengIEEEIEEE Access2169-35362025-01-011313648313650410.1109/ACCESS.2025.359433911105079Optimized Machine Learning-Augmented Hybrid Empirical Models for AlGaN/GaN HEMTs: A Comprehensive AnalysisAhmad Khusro0https://orcid.org/0000-0003-2025-3040Saddam Husain1https://orcid.org/0000-0001-9830-5133Mohammad Hashmi2https://orcid.org/0000-0002-1772-588XDepartment of Electrical and Computer Engineering, School of Engineering and Digital Sciences, Nazarbayev University, Astana, KazakhstanDepartment of Electrical and Computer Engineering, School of Engineering and Digital Sciences, Nazarbayev University, Astana, KazakhstanDepartment of Electrical and Computer Engineering, School of Engineering and Digital Sciences, Nazarbayev University, Astana, KazakhstanPrecise modeling of gallium nitride (GaN) high-electron mobility transistors (HEMTs) is vital for ensuring reliable and scalable RF circuit design, and efficient characterization of the device behavior. This article presents robust hybrid equivalent circuit (EC)–machine learning (ML) frameworks for significantly streamlining the extraction of small-signal model parameters of AlGaN/GaN HEMTs. The extrinsic and intrinsic parameters of the devices are initially extracted using physics-relevant empirical models in Keysight’s advanced design system. Thereafter, six extensively optimized ML regression models, namely decision tree (DT), ensemble learning (EL), support vector regression (SVR), kernel approximation regression (KAR), Gaussian process regression (GPR), and neural networks (NN) are employed to simulate the intrinsic behavior of GaN HEMTs. The models are trained on GaN HEMTs of geometries <inline-formula> <tex-math notation="LaTeX">$4\times 100~\mu $ </tex-math></inline-formula>m, <inline-formula> <tex-math notation="LaTeX">$10\times 220~\mu $ </tex-math></inline-formula>m, and <inline-formula> <tex-math notation="LaTeX">$10\times 250~\mu $ </tex-math></inline-formula>m, while tested on GaN HEMTs of geometries <inline-formula> <tex-math notation="LaTeX">$2\times 200~\mu $ </tex-math></inline-formula>m and <inline-formula> <tex-math notation="LaTeX">$10\times 200~\mu $ </tex-math></inline-formula>m across diverse biasing and frequency conditions. The input features to models include gate–source voltage (VGS), drain–source voltage (VDS), frequency, number of fingers (NF), unity gate width (Wg), and effective gate width (Weff). Finally, a thorough quantitative assessment and detailed comparisons are performed in terms of standard regression tests, mean absolute percentage error, Nash–Sutcliffe efficiency, Kling–Gupta efficiency, training and prediction speed, reliability of model parameters, and simulation agreement with the measured S-parameters. The results demonstrate that among the tested ML models, EL exhibited the lowest mean relative S-parameter errors (2.78–3.75 %), followed by NNs (2.05–6.98 %), DT (2.29–7.35 %), GPR (2.86-8.91 %) SVR (7.83–9.88 %), and KAR (8.26–10.54 %) across diverse GaN HEMTs geometries. This hybrid modeling strategy provides a practical alternative to conventional parameter extraction, offering speed, accuracy, and broader applicability.https://ieeexplore.ieee.org/document/11105079/AlGaN/GaN HEMTdata-driven designhybrid modelingintrinsic parameter extractionmachine learningRF transistor modeling |
| spellingShingle | Ahmad Khusro Saddam Husain Mohammad Hashmi Optimized Machine Learning-Augmented Hybrid Empirical Models for AlGaN/GaN HEMTs: A Comprehensive Analysis IEEE Access AlGaN/GaN HEMT data-driven design hybrid modeling intrinsic parameter extraction machine learning RF transistor modeling |
| title | Optimized Machine Learning-Augmented Hybrid Empirical Models for AlGaN/GaN HEMTs: A Comprehensive Analysis |
| title_full | Optimized Machine Learning-Augmented Hybrid Empirical Models for AlGaN/GaN HEMTs: A Comprehensive Analysis |
| title_fullStr | Optimized Machine Learning-Augmented Hybrid Empirical Models for AlGaN/GaN HEMTs: A Comprehensive Analysis |
| title_full_unstemmed | Optimized Machine Learning-Augmented Hybrid Empirical Models for AlGaN/GaN HEMTs: A Comprehensive Analysis |
| title_short | Optimized Machine Learning-Augmented Hybrid Empirical Models for AlGaN/GaN HEMTs: A Comprehensive Analysis |
| title_sort | optimized machine learning augmented hybrid empirical models for algan gan hemts a comprehensive analysis |
| topic | AlGaN/GaN HEMT data-driven design hybrid modeling intrinsic parameter extraction machine learning RF transistor modeling |
| url | https://ieeexplore.ieee.org/document/11105079/ |
| work_keys_str_mv | AT ahmadkhusro optimizedmachinelearningaugmentedhybridempiricalmodelsforalganganhemtsacomprehensiveanalysis AT saddamhusain optimizedmachinelearningaugmentedhybridempiricalmodelsforalganganhemtsacomprehensiveanalysis AT mohammadhashmi optimizedmachinelearningaugmentedhybridempiricalmodelsforalganganhemtsacomprehensiveanalysis |