Machine learning methods for the nonlinearity mitigation in the physical layer of fiber-optic communication links
| Main Author: | Vladislav Neskorniuk |
|---|---|
| Format: | info Proceeding Journal |
| Bahasa: | eng |
| Terbitan: |
, 2021
|
| Subjects: | |
| Online Access: |
https://zenodo.org/record/5528637 |
Daftar Isi:
- The slides review applications of machine learning to the nonlinearity mitigation in the physical layer of optical telecommunications. The applications are grouped around the three use cases: equalization, pre-distortion, and end-to-end learning. Particular attention is brought to multi-stage neural-network-based nonlinearity equalizers and end-to-end learning of the long-haul high-baudrate fiber-optic communication links.
- This presentation was given as part of the ECOC'21 workshop "How machine learning can revolutionize optical fiber communications?" (https://www.ecoc2021.org/programme/programme-monday#mo1b-ws)