volume | PIER Journals

Abstract

Conventional full-wave methods face prohibitive computational costs for far-field scattering optimization of metasurface-integrated cavity structures. To address this limitation, a lightweight residual neural network is introduced within a two-stage scattering prediction framework. This framework effectively mitigates model degradation. The first stage employs shallow convolutional networks to extract local phase-coupling features. The second stage integrates residual layers with fully connected layers to refine cross-scale scattering responses. A compact CNN-ResNet surrogate model is developed for rapid cavity scattering prediction. With only 2.5×10⁴ parameters and training on 500 full-wave samples spanning 6.0-16.0 GHz, the model achieves high computational efficiency. The proposed approach directly maps binary phase-coded matrices to far-field electromagnetic characteristics. Extensive validation on a cavity structure across 6.0-16.0 GHz demonstrates excellent accuracy. The per-sample runtime is reduced from hours to milliseconds while maintaining prediction errors below 3 dB. These results confirm the effectiveness of the approach in enabling fast and accurate electromagnetic scattering prediction for complex cavity environments. The approach provides a practical solution for metasurface-integrated cavity optimization.

1. Obelleiro-Basteiro, F., J. Luis Rodriguez, and R. J. Burkholder, "An iterative physical optics approach for analyzing the electromagnetic scattering by large open-ended cavities," IEEE Transactions on Antennas and Propagation, Vol. 43, No. 4, 356-361, Apr. 1995.
doi:10.1109/8.376032 Google Scholar

2. Bourlier, Christophe, Hongyang He, Janic Chauveau, Régis Hémon, and Philippe Pouliguen, "RCS of large bent waveguide ducts from a modal analysis combined with the Kirchhoff approximation," Progress In Electromagnetics Research, Vol. 88, 1-38, 2008.
doi:10.2528/pier08101708 Google Scholar

3. Lee, Choon and Shung-Wu Lee, "RCS of a coated circular waveguide terminated by a perfect conductor," IEEE Transactions on Antennas and Propagation, Vol. 35, No. 4, 391-398, Apr. 1987.
doi:10.1109/tap.1987.1144114 Google Scholar

4. Yang, Yunyun, Haoxuan Xin, Yixin Liu, Haoliang Cheng, Yongxing Jin, Chenxia Li, Jianxun Lu, Bo Fang, Zhi Hong, and Xufeng Jing, "Intelligent metasurfaces: Integration of artificial intelligence technology and metasurfaces," Chinese Journal of Physics, Vol. 89, 991-1008, 2024.
doi:10.1016/j.cjph.2024.03.043 Google Scholar

5. Zhao, Tianqi, Bo Fang, Peng Chen, Youhuang Ke, Zhe Kong, Changyu Shen, Chenxia Li, and Xufeng Jing, "High efficiency flexible control of wave beams based on addition and subtraction operations on all dielectric reflection metasurfaces," IEEE Sensors Journal, Vol. 22, No. 5, 4057-4068, Mar. 2022.
doi:10.1109/jsen.2022.3143863 Google Scholar

6. Ma, Qian, Qiang Xiao, Qiao Ru Hong, Xinxin Gao, Vincenzo Galdi, and Tie Jun Cui, "Digital coding metasurfaces: From theory to applications," IEEE Antennas and Propagation Magazine, Vol. 64, No. 4, 96-109, Aug. 2022.
doi:10.1109/map.2022.3169397 Google Scholar

7. Xu, Peng, Wei Xiang Jiang, Xiao Cai, Shi Hao Bai, and Tie Jun Cui, "An integrated coding-metasurface-based array antenna," IEEE Transactions on Antennas and Propagation, Vol. 68, No. 2, 891-899, Feb. 2020.
doi:10.1109/tap.2019.2944529 Google Scholar

8. Chai, Yue, Hui Deng, and Qingxu Xiong, "A dynamically phase tunable metasurface for a broad bandwidth ultra-low radar cross section," IEEE Access, Vol. 8, 53006-53017, 2020.
doi:10.1109/access.2020.2981163 Google Scholar

9. Malkiel, Itzik, Michael Mrejen, Achiya Nagler, Uri Arieli, Lior Wolf, and Haim Suchowski, "Plasmonic nanostructure design and characterization via deep learning," Light: Science & Applications, Vol. 7, No. 1, 60, 2018.
doi:10.1038/s41377-018-0060-7 Google Scholar

10. Paquay, Maurice, Juan-Carlos Iriarte, IÑigo Ederra, Ramon Gonzalo, and Peter de Maagt, "Thin AMC structure for radar cross-section reduction," IEEE Transactions on Antennas and Propagation, Vol. 55, No. 12, 3630-3638, Dec. 2007.
doi:10.1109/tap.2007.910306 Google Scholar

11. Ameri, Edris, Seyed Hassan Esmaeli, and Seyed Hassan Sedighy, "Ultra wideband radar cross section reduction by using polarization conversion metasurfaces," Scientific Reports, Vol. 9, No. 1, 478, Jan. 2019.
doi:10.1038/s41598-018-36542-6 Google Scholar

12. Pan, Yibo, Feng Lan, Yaxin Zhang, Hongxin Zeng, Luyang Wang, Tianyang Song, Guiju He, and Ziqiang Yang, "Dual-band multifunctional coding metasurface with a mingled anisotropic aperture for polarized manipulation in full space," Photonics Research, Vol. 10, No. 2, 416-425, 2022.
doi:10.1364/prj.444773 Google Scholar

13. Amboli, Jayeeta, Bruno Gallas, Guillaume Demésy, and Nicolas Bonod, "Design and analysis of chiral and achiral metasurfaces with the finite element method," Optics Express, Vol. 31, No. 26, 43147-43162, 2023.
doi:10.1364/oe.500540 Google Scholar

14. Jia, Xiao, Fan Yang, Yinghong Wen, Maokun Li, and Shenheng Xu, "Characteristic model and efficient FDTD-SPM algorithm for fishnet metasurfaces analysis," IEEE Transactions on Antennas and Propagation, Vol. 70, No. 10, 8729-8738, Oct. 2022.
doi:10.1109/tap.2022.3177480 Google Scholar

15. Teng, Yan, Chun Li, Shaochen Li, Yuhua Xiao, and Ling Jiang, "Efficient design method for terahertz broadband metasurface patterns via deep learning," Optics & Laser Technology, Vol. 160, 109058, 2023.
doi:10.1016/j.optlastec.2022.109058 Google Scholar

16. Fu, Jiahui, Zhihu Yang, Meng Liu, Huiyun Zhang, and Yuping Zhang, "Highly-efficient design method for coding metasurfaces based on deep learning," Optics Communications, Vol. 529, 129043, 2023.
doi:10.1016/j.optcom.2022.129043 Google Scholar

17. Donda, Krupali, Yifan Zhu, Aurélien Merkel, Shi-Wang Fan, Liyun Cao, Sheng Wan, and Badreddine Assouar, "Ultrathin acoustic absorbing metasurface based on deep learning approach," Smart Materials and Structures, Vol. 30, No. 8, 085003, 2021.
doi:10.1088/1361-665x/ac0675 Google Scholar

18. He, Qingting, Haiyan Chen, Qian Liu, Xin Yao, Fengxia Li, Difei Liang, Jianliang Xie, and Longjiang Deng, "Ultra-wideband and wide-angle RCS reduction of a concave structure based on a chessboard polarization conversion metasurfaces," Journal of Physics D: Applied Physics, Vol. 57, No. 3, 035104, 2023.
doi:10.1088/1361-6463/ad005e Google Scholar

19. He, Qingting, Haiyan Chen, Qian Liu, Xin Yao, Fengxia Li, Difei Liang, Jianliang Xie, and Longjiang Deng, "Design of broadband and ultra-wide-angle low-RCS open-ended cavity based on phase cancellation," International Journal of RF and Microwave Computer-Aided Engineering, Vol. 2023, No. 1, 9958074, 2023.
doi:10.1155/2023/9958074 Google Scholar

20. Zunair, Hasib and A. Ben Hamza, "Sharp U-Net: Depthwise convolutional network for biomedical image segmentation," Computers in Biology and Medicine, Vol. 136, 104699, 2021.
doi:10.1016/j.compbiomed.2021.104699 Google Scholar

21. Liu, Ze, Yutong Lin, Yue Cao, Han Hu, Yixuan Wei, Zheng Zhang, Stephen Lin, and Baining Guo, "Swin transformer: Hierarchical vision transformer using shifted windows," 2021 IEEE/CVF International Conference on Computer Vision (ICCV), 9992-10002, Montreal, QC, Canada, 2021.
doi:10.1109/ICCV48922.2021.00986

22. Huang, Gao, Zhuang Liu, Laurens Van Der Maaten, and Kilian Q. Weinberger, "Densely connected convolutional networks," Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 4700-4708, 2017.

Dr. Xi Liu

Dr. Peng Nian

Dr. Yu Zhang

Professor Yi Ren

Dr. Yi-Xin Guo

Dr. Yang-Chun Gao

Dr. Bing Chen