Deep Learning Designs of Anisotropic Metasurfaces in Ultrawideband Based on Generative Adversarial Networks
Metasurfaces have been developed as a promising approach for manipulating electromagnetic waves. Recently, deep learning algorithms have been introduced to design metasurfaces, but the network can only output one solution for each desired input and suffers from nonunique issue. To overcome the afore...
| Main Authors: | , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Wiley
2020-09-01
|
| Series: | Advanced Intelligent Systems |
| Subjects: | |
| Online Access: | https://doi.org/10.1002/aisy.202000068 |
| _version_ | 1830482148686036992 |
|---|---|
| author | Hai Peng Wang Yun Bo Li He Li Shu Yue Dong Che Liu Shi Jin Tie Jun Cui |
| author_facet | Hai Peng Wang Yun Bo Li He Li Shu Yue Dong Che Liu Shi Jin Tie Jun Cui |
| author_sort | Hai Peng Wang |
| collection | DOAJ |
| description | Metasurfaces have been developed as a promising approach for manipulating electromagnetic waves. Recently, deep learning algorithms have been introduced to design metasurfaces, but the network can only output one solution for each desired input and suffers from nonunique issue. To overcome the aforementioned challenges, a deep neural network model for inverse designs of anisotropic metasurfaces with full phase properties in ultrawideband is proposed. Given the target reflection spectra as inputs, the candidate metasurface patterns are generated through a generative adversarial network (GAN), and the corresponding predictions are simply achieved by the accurate forward neural network model to match the target spectra in the whole band with high fidelity. By training the generator and discriminator in GAN in an alternating order combined with setting a threshold of discriminator loss to trigger the phase prediction, the proposed method is much more efficient and consumes less time in the training process. Numerical simulations and experimental results demonstrate that the reflection phases of the generated meta‐atoms have excellent agreements with the given targets, providing an efficient way in automatically designing metasurfaces. The most important advantage of this approach over the previous schemes is to improve the design speed significantly with very good accuracy. |
| first_indexed | 2024-12-21T17:29:46Z |
| format | Article |
| id | doaj.art-f506904424de4aa8b2439196428c3de7 |
| institution | Directory Open Access Journal |
| issn | 2640-4567 |
| language | English |
| last_indexed | 2024-12-21T17:29:46Z |
| publishDate | 2020-09-01 |
| publisher | Wiley |
| record_format | Article |
| series | Advanced Intelligent Systems |
| spelling | doaj.art-f506904424de4aa8b2439196428c3de72022-12-21T18:55:57ZengWileyAdvanced Intelligent Systems2640-45672020-09-0129n/an/a10.1002/aisy.202000068Deep Learning Designs of Anisotropic Metasurfaces in Ultrawideband Based on Generative Adversarial NetworksHai Peng Wang0Yun Bo Li1He Li2Shu Yue Dong3Che Liu4Shi Jin5Tie Jun Cui6State Key Laboratory of Millimeter Waves Southeast University Nanjing 210096 ChinaState Key Laboratory of Millimeter Waves Southeast University Nanjing 210096 ChinaState Key Laboratory of Millimeter Waves Southeast University Nanjing 210096 ChinaState Key Laboratory of Millimeter Waves Southeast University Nanjing 210096 ChinaState Key Laboratory of Millimeter Waves Southeast University Nanjing 210096 ChinaNational Mobile Communications Research Laboratory Southeast University Nanjing 210096 ChinaState Key Laboratory of Millimeter Waves Southeast University Nanjing 210096 ChinaMetasurfaces have been developed as a promising approach for manipulating electromagnetic waves. Recently, deep learning algorithms have been introduced to design metasurfaces, but the network can only output one solution for each desired input and suffers from nonunique issue. To overcome the aforementioned challenges, a deep neural network model for inverse designs of anisotropic metasurfaces with full phase properties in ultrawideband is proposed. Given the target reflection spectra as inputs, the candidate metasurface patterns are generated through a generative adversarial network (GAN), and the corresponding predictions are simply achieved by the accurate forward neural network model to match the target spectra in the whole band with high fidelity. By training the generator and discriminator in GAN in an alternating order combined with setting a threshold of discriminator loss to trigger the phase prediction, the proposed method is much more efficient and consumes less time in the training process. Numerical simulations and experimental results demonstrate that the reflection phases of the generated meta‐atoms have excellent agreements with the given targets, providing an efficient way in automatically designing metasurfaces. The most important advantage of this approach over the previous schemes is to improve the design speed significantly with very good accuracy.https://doi.org/10.1002/aisy.202000068anisotropic reflective phase responsegenerative adversarial networksinverse designsmetasurfacesultrawideband |
| spellingShingle | Hai Peng Wang Yun Bo Li He Li Shu Yue Dong Che Liu Shi Jin Tie Jun Cui Deep Learning Designs of Anisotropic Metasurfaces in Ultrawideband Based on Generative Adversarial Networks Advanced Intelligent Systems anisotropic reflective phase response generative adversarial networks inverse designs metasurfaces ultrawideband |
| title | Deep Learning Designs of Anisotropic Metasurfaces in Ultrawideband Based on Generative Adversarial Networks |
| title_full | Deep Learning Designs of Anisotropic Metasurfaces in Ultrawideband Based on Generative Adversarial Networks |
| title_fullStr | Deep Learning Designs of Anisotropic Metasurfaces in Ultrawideband Based on Generative Adversarial Networks |
| title_full_unstemmed | Deep Learning Designs of Anisotropic Metasurfaces in Ultrawideband Based on Generative Adversarial Networks |
| title_short | Deep Learning Designs of Anisotropic Metasurfaces in Ultrawideband Based on Generative Adversarial Networks |
| title_sort | deep learning designs of anisotropic metasurfaces in ultrawideband based on generative adversarial networks |
| topic | anisotropic reflective phase response generative adversarial networks inverse designs metasurfaces ultrawideband |
| url | https://doi.org/10.1002/aisy.202000068 |
| work_keys_str_mv | AT haipengwang deeplearningdesignsofanisotropicmetasurfacesinultrawidebandbasedongenerativeadversarialnetworks AT yunboli deeplearningdesignsofanisotropicmetasurfacesinultrawidebandbasedongenerativeadversarialnetworks AT heli deeplearningdesignsofanisotropicmetasurfacesinultrawidebandbasedongenerativeadversarialnetworks AT shuyuedong deeplearningdesignsofanisotropicmetasurfacesinultrawidebandbasedongenerativeadversarialnetworks AT cheliu deeplearningdesignsofanisotropicmetasurfacesinultrawidebandbasedongenerativeadversarialnetworks AT shijin deeplearningdesignsofanisotropicmetasurfacesinultrawidebandbasedongenerativeadversarialnetworks AT tiejuncui deeplearningdesignsofanisotropicmetasurfacesinultrawidebandbasedongenerativeadversarialnetworks |