Advanced search
2 files | 6.39 MB Add to list

Spectral bayesian optimization using a physics-informed rational Szego kernel for microwave design

Yens Lindemans (UGent) , Thijs Ullrick (UGent) , Ivo Couckuyt (UGent) , Tim Pattyn (UGent) , Dirk Deschrijver (UGent) , Dries Vande Ginste (UGent) and Tom Dhaene (UGent)
Author
Organization
Project
Abstract
Microwave device design increasingly relies on surrogate modeling to accelerate optimization and reduce costly electromagnetic (EM) simulations. This article presents a spectral Bayesian optimization (SBO) framework leveraging a physics-informed Gaussian process (GP) with a rational complex-valued Szeg & ouml; kernel and input warping to enhance surrogate accuracy and data efficiency. Unlike conventional methods that model scalar objectives, our approach directly learns the complex-valued frequency response, enforcing causality and Hermitian symmetry. Effectiveness is demonstrated in two cases: a zig-zag microstrip bandpass filter optimized for magnitude response and a passive differential equalizer optimized for both transmission magnitude and group delay. By embedding prior physics and modeling directly in the frequency domain, the method enables accurate, sample-efficient optimization of frequency-dependent behavior. This work shows how physics-informed Bayesian optimization (BO) can significantly improve microwave device design efficiency.
Keywords
Kernel, Optimization, Scattering parameters, Gaussian processes, Frequency response, Bayes methods, Covariance matrices, Vectors, Training, Microwave filters, Bayesian optimization (BO), Gaussian process (GP), machine learning (ML), microwave devices, physics-informed, Szeg & ouml, kernel

Downloads

  • (...).pdf
    • full text (Published version)
    • |
    • UGent only
    • |
    • PDF
    • |
    • 3.71 MB
  • 8885 acc.pdf
    • full text (Accepted manuscript)
    • |
    • open access
    • |
    • PDF
    • |
    • 2.68 MB

Citation

Please use this url to cite or link to this publication:

MLA
Lindemans, Yens, et al. “Spectral Bayesian Optimization Using a Physics-Informed Rational Szego Kernel for Microwave Design.” IEEE TRANSACTIONS ON COMPONENTS PACKAGING AND MANUFACTURING TECHNOLOGY, vol. 15, no. 9, 2025, pp. 1836–46, doi:10.1109/TCPMT.2025.3592441.
APA
Lindemans, Y., Ullrick, T., Couckuyt, I., Pattyn, T., Deschrijver, D., Vande Ginste, D., & Dhaene, T. (2025). Spectral bayesian optimization using a physics-informed rational Szego kernel for microwave design. IEEE TRANSACTIONS ON COMPONENTS PACKAGING AND MANUFACTURING TECHNOLOGY, 15(9), 1836–1846. https://doi.org/10.1109/TCPMT.2025.3592441
Chicago author-date
Lindemans, Yens, Thijs Ullrick, Ivo Couckuyt, Tim Pattyn, Dirk Deschrijver, Dries Vande Ginste, and Tom Dhaene. 2025. “Spectral Bayesian Optimization Using a Physics-Informed Rational Szego Kernel for Microwave Design.” IEEE TRANSACTIONS ON COMPONENTS PACKAGING AND MANUFACTURING TECHNOLOGY 15 (9): 1836–46. https://doi.org/10.1109/TCPMT.2025.3592441.
Chicago author-date (all authors)
Lindemans, Yens, Thijs Ullrick, Ivo Couckuyt, Tim Pattyn, Dirk Deschrijver, Dries Vande Ginste, and Tom Dhaene. 2025. “Spectral Bayesian Optimization Using a Physics-Informed Rational Szego Kernel for Microwave Design.” IEEE TRANSACTIONS ON COMPONENTS PACKAGING AND MANUFACTURING TECHNOLOGY 15 (9): 1836–1846. doi:10.1109/TCPMT.2025.3592441.
Vancouver
1.
Lindemans Y, Ullrick T, Couckuyt I, Pattyn T, Deschrijver D, Vande Ginste D, et al. Spectral bayesian optimization using a physics-informed rational Szego kernel for microwave design. IEEE TRANSACTIONS ON COMPONENTS PACKAGING AND MANUFACTURING TECHNOLOGY. 2025;15(9):1836–46.
IEEE
[1]
Y. Lindemans et al., “Spectral bayesian optimization using a physics-informed rational Szego kernel for microwave design,” IEEE TRANSACTIONS ON COMPONENTS PACKAGING AND MANUFACTURING TECHNOLOGY, vol. 15, no. 9, pp. 1836–1846, 2025.
@article{01K6FAGNKQNN498G62G1XPHARC,
  abstract     = {{Microwave device design increasingly relies on surrogate modeling to accelerate optimization and reduce costly electromagnetic (EM) simulations. This article presents a spectral Bayesian optimization (SBO) framework leveraging a physics-informed Gaussian process (GP) with a rational complex-valued Szeg & ouml; kernel and input warping to enhance surrogate accuracy and data efficiency. Unlike conventional methods that model scalar objectives, our approach directly learns the complex-valued frequency response, enforcing causality and Hermitian symmetry. Effectiveness is demonstrated in two cases: a zig-zag microstrip bandpass filter optimized for magnitude response and a passive differential equalizer optimized for both transmission magnitude and group delay. By embedding prior physics and modeling directly in the frequency domain, the method enables accurate, sample-efficient optimization of frequency-dependent behavior. This work shows how physics-informed Bayesian optimization (BO) can significantly improve microwave device design efficiency.}},
  author       = {{Lindemans, Yens and Ullrick, Thijs and Couckuyt, Ivo and Pattyn, Tim and Deschrijver, Dirk and Vande Ginste, Dries and Dhaene, Tom}},
  issn         = {{2156-3950}},
  journal      = {{IEEE TRANSACTIONS ON COMPONENTS PACKAGING AND MANUFACTURING TECHNOLOGY}},
  keywords     = {{Kernel,Optimization,Scattering parameters,Gaussian processes,Frequency response,Bayes methods,Covariance matrices,Vectors,Training,Microwave filters,Bayesian optimization (BO),Gaussian process (GP),machine learning (ML),microwave devices,physics-informed,Szeg & ouml,kernel}},
  language     = {{eng}},
  number       = {{9}},
  pages        = {{1836--1846}},
  title        = {{Spectral bayesian optimization using a physics-informed rational Szego kernel for microwave design}},
  url          = {{http://doi.org/10.1109/TCPMT.2025.3592441}},
  volume       = {{15}},
  year         = {{2025}},
}

Altmetric
View in Altmetric
Web of Science
Times cited: