A modulation format recognition and optical signal-to-noise ratio monitoring scheme based on residual network and Taylor score pruning
These videos have been matched automatically. Contact us if they are not relevant.
These videos have been matched automatically. Contact us if they are not relevant.
View abstract on PubMed
Summary
This summary is machine-generated.A novel SA-ResNet model efficiently monitors modulation formats and optical signal-to-noise ratio in coherent optical systems. This lightweight approach offers high accuracy for advanced optical networks.
Area Of Science
- Optical Communications Engineering
- Machine Learning for Signal Processing
- Telecommunications Network Monitoring
Background
- Real-time monitoring of modulation formats (MF) and optical signal-to-noise ratio (OSNR) is crucial for dynamic optical networks.
- Existing methods may lack efficiency or accuracy for complex modulation schemes in coherent systems.
Purpose Of The Study
- To develop an efficient joint monitoring method for MF and OSNR in coherent optical communication systems.
- To apply a lightweight deep learning model suitable for resource-constrained optical fiber monitoring.
Main Methods
- Proposed a residual network with an attention mechanism (SA-ResNet) for joint MF and OSNR monitoring.
- Evaluated the model on various M-ary Quadrature Amplitude Modulation (MQAM) signals (QPSK to 128QAM).
- Utilized Taylor pruning to reduce computational complexity (FLOPs and parameter memory).
Main Results
- Achieved 100% MF recognition accuracy and an average absolute OSNR error of 0.34 dB post-pruning and fine-tuning.
- 5-fold cross-validation yielded 99.988% MF accuracy and 0.32 dB average OSNR error.
- Reduced FLOPs from 40.5 M to 9.5 M and memory from 2.6 M to 0.5 M.
Conclusions
- The SA-ResNet model demonstrates high accuracy and efficiency for joint MF and OSNR monitoring.
- The model's low computational resource requirement makes it suitable for lightweight optical fiber monitoring systems.
- This approach supports the advancement of future dynamic and heterogeneous optical networks.
These videos have been matched automatically. Contact us if they are not relevant.
These videos have been matched automatically. Contact us if they are not relevant.
