Real-Time Mode Decomposition in Multimode Fiber Communications Using Optical Diffractive Neural Network
Publikation: Beitrag in Buch/Konferenzbericht/Sammelband/Gutachten › Beitrag in Konferenzband › Beigetragen › Begutachtung
Beitragende
Abstract
We propose a novel approach for real-time decomposition of multimode fibers using optical diffractive neural networks. Leveraging the physics-informed nature of the model, training using pure eigenmode data enables precise mode decomposition, which is of great importance for space division multiplexing in multimode fiber communication.
Details
| Originalsprache | Englisch |
|---|---|
| Titel | 2025 IEEE Photonics Conference, IPC 2025 - Proceedings |
| Herausgeber (Verlag) | Institute of Electrical and Electronics Engineers (IEEE) |
| Seiten | 1-2 |
| ISBN (elektronisch) | 979-8-3315-2559-0 |
| ISBN (Print) | 979-8-3315-2560-6 |
| Publikationsstatus | Veröffentlicht - Dez. 2025 |
| Peer-Review-Status | Ja |
Publikationsreihe
| Reihe | IEEE Photonics Conference (IPC) |
|---|---|
| ISSN | 2374-0140 |
Konferenz
| Titel | 2025 IEEE Photonics Conference |
|---|---|
| Kurztitel | IPC 2025 |
| Dauer | 9 - 13 November 2025 |
| Webseite | |
| Ort | Resorts World Sentosa |
| Stadt | Singapore |
| Land | Singapur |
Schlagworte
ASJC Scopus Sachgebiete
- Computernetzwerke und -kommunikation - deep learning, diffractive neural network, mode decomposition, multimode fiber, optical computing, physics-driven
- Hardware und Architektur - deep learning, diffractive neural network, mode decomposition, multimode fiber, optical computing, physics-driven
- Signalverarbeitung - deep learning, diffractive neural network, mode decomposition, multimode fiber, optical computing, physics-driven
- Elektrotechnik und Elektronik - deep learning, diffractive neural network, mode decomposition, multimode fiber, optical computing, physics-driven
- Elektronische, optische und magnetische Materialien - deep learning, diffractive neural network, mode decomposition, multimode fiber, optical computing, physics-driven
- Atom- und Molekularphysik sowie Optik - deep learning, diffractive neural network, mode decomposition, multimode fiber, optical computing, physics-driven
Schlagwörter
- deep learning, diffractive neural network, mode decomposition, multimode fiber, optical computing, physics-driven