Globalized Simulation-Driven Miniaturization of Microwave Circuits by Means of Dimensionality-Reduced Constrained Surrogates
Abstract
Small size has become a crucial prerequisite in the design of modern microwave components. Miniaturized devices are essential for a number of application areas, including wireless communications, 5G/6G technology, wearable devices, or the internet of things. Notwithstanding, size reduction generally degrades the electrical performance of microwave systems. Therefore, trade-off solutions have to be sought that represent acceptable compromises between the ability to meet the design targets and physical compactness. From an optimization perspective, this poses a constrained task, which is computationally expensive because a reliable evaluation of microwave components has to rely on full-wave electromagnetic (EM) analysis. Furthermore, due to its constrained nature, size reduction is a multimodal problem, i.e., the results are highly dependent on the initial design. Thus, utilization of global search algorithms is advisable in principle, yet, often undoable in practice because of the associated computational expenses, especially when using nature-inspired procedures. This paper introduces a novel technique for globalized miniaturization of microwave components. Our technique starts by identifying the feasible region boundary, and by constructing a dimensionality-reduced surrogate model therein. Global optimization of the metamodel is followed by EM-driven local tuning. Application of the domain-confined surrogate ensures low cost of the entire procedure, further reduced by the incorporation of variable-fidelity EM simulations. Our framework is validated using two microstrip couplers, and compared to nature-inspired optimization, as well as gradient-based size reduction. The results indicate superior miniaturization rates and low running cost, which make the presented algorithm a potential candidate for efficient simulation-based design of compact structures.
Citations
-
6
CrossRef
-
0
Web of Science
-
6
Scopus
Authors (3)
Cite as
Full text
- Publication version
- Accepted or Published Version
- DOI:
- Digital Object Identifier (open in new tab) 10.1038/s41598-022-20728-0
- License
- open in new tab
Keywords
Details
- Category:
- Articles
- Type:
- artykuły w czasopismach
- Published in:
-
Scientific Reports
no. 12,
ISSN: 2045-2322 - Language:
- English
- Publication year:
- 2022
- Bibliographic description:
- Kozieł S., Pietrenko-Dąbrowska A., Mahrokh M.: Globalized Simulation-Driven Miniaturization of Microwave Circuits by Means of Dimensionality-Reduced Constrained Surrogates// Scientific Reports -Vol. 12,iss. 1 (2022), s.16418-
- DOI:
- Digital Object Identifier (open in new tab) 10.1038/s41598-022-20728-0
- Sources of funding:
-
- Statutory activity/subsidy
- Verified by:
- Gdańsk University of Technology
seen 97 times