Optical Neural Network in Free-Space and Nanophotonics
The explosive data growth has resulted in increased computing costs. As Moore’s Law is increasingly slowing down, the traditional computing approach based on the von Neumann architecture is gradually becoming unable to fulfill future computing needs. However, optical neural networks have...
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2023-01-01
|
| Series: | IEEE Access |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/10197423/ |
| _version_ | 1797736958326210560 |
|---|---|
| author | Zhenlin Sun Miao Yu Zhengxun Song Weiwen Liu Gangyao Xing Muhan Zhou |
| author_facet | Zhenlin Sun Miao Yu Zhengxun Song Weiwen Liu Gangyao Xing Muhan Zhou |
| author_sort | Zhenlin Sun |
| collection | DOAJ |
| description | The explosive data growth has resulted in increased computing costs. As Moore’s Law is increasingly slowing down, the traditional computing approach based on the von Neumann architecture is gradually becoming unable to fulfill future computing needs. However, optical neural networks have emerged as a potential solution because of their high speed, high bandwidth, and capability to subdue the bottleneck problem of computing power. With the development of optics and nanophotonics, it is possible to implement complex optical neural networks in free-space and nanophotonic platforms. This article reviews the research progress of optical neural networks. Firstly, various methods of implementing optical matrix calculations are described. Secondly, the construction method of optical neural network in free space and nanophotonic platform is introduced respectively. In free space, based on 4f system and diffractive optical elements and in nanophotonic, relied on optical waveguide devices such as microring resonator or Mach-Zehnder Interferometer. Thirdly, we introduce the methods in training and nonlinear activity. Finally, we summarized the current research status and challenges of optical neural networks. In the future, optical neural networks have great application value. |
| first_indexed | 2024-03-12T13:21:27Z |
| format | Article |
| id | doaj.art-75e0fa1f661c4f95bc1b5fca4a7ca900 |
| institution | Directory Open Access Journal |
| issn | 2169-3536 |
| language | English |
| last_indexed | 2024-03-12T13:21:27Z |
| publishDate | 2023-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Access |
| spelling | doaj.art-75e0fa1f661c4f95bc1b5fca4a7ca9002023-08-25T23:00:55ZengIEEEIEEE Access2169-35362023-01-0111886568866910.1109/ACCESS.2023.330023110197423Optical Neural Network in Free-Space and NanophotonicsZhenlin Sun0https://orcid.org/0009-0009-3073-8958Miao Yu1Zhengxun Song2Weiwen Liu3Gangyao Xing4Muhan Zhou5International Research Centre for Nano Handling and Manufacturing of China, Changchun University of Science and Technology, Changchun, ChinaInternational Research Centre for Nano Handling and Manufacturing of China, Changchun University of Science and Technology, Changchun, ChinaInternational Research Centre for Nano Handling and Manufacturing of China, Changchun University of Science and Technology, Changchun, ChinaInternational Research Centre for Nano Handling and Manufacturing of China, Changchun University of Science and Technology, Changchun, ChinaInternational Research Centre for Nano Handling and Manufacturing of China, Changchun University of Science and Technology, Changchun, ChinaInternational Research Centre for Nano Handling and Manufacturing of China, Changchun University of Science and Technology, Changchun, ChinaThe explosive data growth has resulted in increased computing costs. As Moore’s Law is increasingly slowing down, the traditional computing approach based on the von Neumann architecture is gradually becoming unable to fulfill future computing needs. However, optical neural networks have emerged as a potential solution because of their high speed, high bandwidth, and capability to subdue the bottleneck problem of computing power. With the development of optics and nanophotonics, it is possible to implement complex optical neural networks in free-space and nanophotonic platforms. This article reviews the research progress of optical neural networks. Firstly, various methods of implementing optical matrix calculations are described. Secondly, the construction method of optical neural network in free space and nanophotonic platform is introduced respectively. In free space, based on 4f system and diffractive optical elements and in nanophotonic, relied on optical waveguide devices such as microring resonator or Mach-Zehnder Interferometer. Thirdly, we introduce the methods in training and nonlinear activity. Finally, we summarized the current research status and challenges of optical neural networks. In the future, optical neural networks have great application value.https://ieeexplore.ieee.org/document/10197423/Optical neural networkoptical matrix computingnanophotonicFourier transformdiffractive opticsoptical waveguide |
| spellingShingle | Zhenlin Sun Miao Yu Zhengxun Song Weiwen Liu Gangyao Xing Muhan Zhou Optical Neural Network in Free-Space and Nanophotonics IEEE Access Optical neural network optical matrix computing nanophotonic Fourier transform diffractive optics optical waveguide |
| title | Optical Neural Network in Free-Space and Nanophotonics |
| title_full | Optical Neural Network in Free-Space and Nanophotonics |
| title_fullStr | Optical Neural Network in Free-Space and Nanophotonics |
| title_full_unstemmed | Optical Neural Network in Free-Space and Nanophotonics |
| title_short | Optical Neural Network in Free-Space and Nanophotonics |
| title_sort | optical neural network in free space and nanophotonics |
| topic | Optical neural network optical matrix computing nanophotonic Fourier transform diffractive optics optical waveguide |
| url | https://ieeexplore.ieee.org/document/10197423/ |
| work_keys_str_mv | AT zhenlinsun opticalneuralnetworkinfreespaceandnanophotonics AT miaoyu opticalneuralnetworkinfreespaceandnanophotonics AT zhengxunsong opticalneuralnetworkinfreespaceandnanophotonics AT weiwenliu opticalneuralnetworkinfreespaceandnanophotonics AT gangyaoxing opticalneuralnetworkinfreespaceandnanophotonics AT muhanzhou opticalneuralnetworkinfreespaceandnanophotonics |