Multi-Bit Phase-Coded Metamaterial With Honeycomb EM Structure for Scattering Manipulation Using Deep Learning-Based Method

A deep-learning inverse design method for generating multi-bit phase-coded metamaterials is proposed, aiming at broadband electromagnetic (EM) scattering manipulation with stable performance under oblique incidence. First, a deep learning-based approach for broadband phase gradient manipulation is analyzed using an equivalent circuit model (ECM), resulting in a bandwidth of 8.0–18.0 GHz. Then, to enhance performance under oblique incidence, a miniaturized folded tightly coupled dipole structure is proposed to suppress parasitic resonance. Furthermore, a design method combining deep learning with intelligent optimization algorithms is investigated to generate and optimize the dipole, ensuring stable operation at incident angles up to 45°. By leveraging ECM principles and previous EM knowledge, the physical dimensions of the unit cell generated by the deep-learning model are reduced from five to four parameters, and five-bit phase-coded unit cells are generated accordingly. Finally, a novel phase-coded metamaterial based on a honeycomb EM structure consisting of <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$29\times 33$</tex-math> </inline-formula> unit cells, with a thickness of <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$0.36\lambda _{C}$</tex-math> </inline-formula> and a periodicity of 10.3 mm, is designed and simulated for radar cross-section (RCS) reduction under linear polarization (LP) incident waves. For demonstration, one prototype is fabricated and measured. The results indicate a broadband phase gradient from 8.0 to 18.0 GHz (76.9%) up to an incident angle of 45° in the case of dual LP. Regarding RCS reduction, the scattering far-field is reduced by more than 20 dBsm under LP normal incidence, with a 10 dBsm reduction maintained even at a 45° oblique incidence.

Authors

Institutions

Publication Details

Journal
IEEE Transactions on Microwave Theory and Techniques
Published
2025-06-09
DOI
https://doi.org/10.1109/tmtt.2025.3574674
Citations
1
Primary Topic
Advanced Antenna and Metasurface Technologies
Type
article
Field-Weighted Citation Impact
4.99

Funders

Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
article

Multi-Bit Phase-Coded Metamaterial With Honeycomb EM Structure for Scattering Manipulation Using Deep Learning-Based Method

Kun Tang, Wenjie Feng, Jiawei Liu, Xin Xiu et al.
1 citations
IEEE Transactions on Microwave Theory and Techniques
Advanced Antenna and Metasurface Technologies
4.99
article

Multi-Bit Phase-Coded Metamaterial With Honeycomb EM Structure for Scattering Manipulation Using Deep Learning-Based Method

Kun Tang, Wenjie Feng, Jiawei Liu, Xin Xiu, Wenquan Che, Weizhi Chen
article en
1 citations

Abstract

A deep-learning inverse design method for generating multi-bit phase-coded metamaterials is proposed, aiming at broadband electromagnetic (EM) scattering manipulation with stable performance under oblique incidence. First, a deep learning-based approach for broadband phase gradient manipulation is analyzed using an equivalent circuit model (ECM), resulting in a bandwidth of 8.0–18.0 GHz. Then, to enhance performance under oblique incidence, a miniaturized folded tightly coupled dipole structure is proposed to suppress parasitic resonance. Furthermore, a design method combining deep learning with intelligent optimization algorithms is investigated to generate and optimize the dipole, ensuring stable operation at incident angles up to 45°. By leveraging ECM principles and previous EM knowledge, the physical dimensions of the unit cell generated by the deep-learning model are reduced from five to four parameters, and five-bit phase-coded unit cells are generated accordingly. Finally, a novel phase-coded metamaterial based on a honeycomb EM structure consisting of <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$29\times 33$</tex-math> </inline-formula> unit cells, with a thickness of <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$0.36\lambda _{C}$</tex-math> </inline-formula> and a periodicity of 10.3 mm, is designed and simulated for radar cross-section (RCS) reduction under linear polarization (LP) incident waves. For demonstration, one prototype is fabricated and measured. The results indicate a broadband phase gradient from 8.0 to 18.0 GHz (76.9%) up to an incident angle of 45° in the case of dual LP. Regarding RCS reduction, the scattering far-field is reduced by more than 20 dBsm under LP normal incidence, with a 10 dBsm reduction maintained even at a 45° oblique incidence.

IEEE Transactions on Microwave Theory and TechniquesVol. 73(10)
South China Normal University (CN), MWT Materials (United States) (US), South China University of Technology (CN)
National Natural Science Foundation of China
Openalex Percentile: Top 6%
Advanced Antenna and Metasurface Technologies
4.99
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.