Convolutional Neural Networks for Optical Performance Monitoring
2019
We demonstrate accurate OSNR estimation using convolutional neural networks based on the constellation image of a QAM signal. Simulation and experiment results demonstrate estimation error within 0.1 dB. Neural network training on simulated and successful testing on experimental data was also demonstrated.
Keywords:
- Correction
- Source
- Cite
- Save
- Machine Reading By IdeaReader
0
References
3
Citations
NaN
KQI