Flexible and Differentiable Optical Fiber Channel Modeling Based on a Neural Network Module
Optical fiber channel is emulated flexibly and accurately by a neural network module which is differentiable naturally and has good generalization over GCS and PCS modulation format with little normalized MSE (<=0.025).
Authors
- Tao Jia (ORCID: https://orcid.org/0000-0003-2802-447X)
- Zhen Zou (ORCID: https://orcid.org/0000-0002-1651-591X)
- Rui Jiang (ORCID: https://orcid.org/0000-0002-6275-1972)
- Kaihong Wang
- Zhi Wang (ORCID: https://orcid.org/0000-0002-0293-4099)
- Yang Li (ORCID: https://orcid.org/0000-0002-4342-3121)
- Huiying Wang (ORCID: https://orcid.org/0000-0002-9798-4258)
- Chao Shang (ORCID: https://orcid.org/0000-0002-5834-9796)
- Z. L. Xie (ORCID: https://orcid.org/0009-0009-7960-4360)
- Xin Ding
- Ningxin Jiao
Institutions
- Beijing Institute of Technology (CN)
- Beijing Jiaotong University (CN)
Publication Details
- Published
- 2024-11-02
- DOI
- https://doi.org/10.1109/acp/ipoc63121.2024.10809675
- Primary Topic
- Semiconductor Lasers and Optical Devices
- Type
- article
- Field-Weighted Citation Impact
- 0.00