Fast signal quality monitoring for coherent communications enabled by CNN-based EVM estimation

---

Published on Journal of Optical Communications and Networking (JOCN) Special Issue on Machine Learning Applied to QoT Estimation in Optical Networks

Links to the manuscript:

• Chalmers (open access authors' version): https://research.chalmers.se/en/publication/521965
• IEEEXplore: https://ieeexplore.ieee.org/document/9326316
• OSA: https://www.osapublishing.org/jocn/abstract.cfm?uri=jocn-13-4-B12

The dataset is available under DOI 10.21227/1684-a275.

---

Authors: Yuchuan Fan, Aleksejs Udalcovs, Xiaodan Pang, Carlos Natalino, Marija Furdek, Sergei Popov, Oskars Ozolins

---

Abstract: We propose a fast and accurate signal quality monitoring scheme that uses convolutional neural networks (CNN) for error vector magnitude (EVM) estimation in coherent optical communications. We build a regression model to extract EVM information from complex signal constellation diagrams using a small number of received symbols. For the additive white Gaussian noise (AWGN) impaired channel, the proposed EVM estimation scheme shows a normalized mean absolute estimation error of 3.7% for quadrature phase shift keying (QPSK), 2.2% for 16-ary quadrature amplitude modulation (16QAM), and 1.1% for 64QAM signals, requiring only 100 symbols per constellation cluster in each observation period. Therefore, it can be used as a low-complexity alternative to conventional bit-error-rate (BER) estimation, enabling solutions for intelligent optical performance monitoring.

Test the implementation of this work in Google Colab:

Link to the Google Colab file: https://colab.research.google.com/drive/1B7KkZAsFtUaBqsdPGQhYwxCLeYPVAdgr?usp=sharing

Citing this work

Paper:

@ARTICLE{FanEtAl:JOCN:EVM:2021,
  author={Y. {Fan} and A. {Udalcovs} and X. {Pang} and C. {Natalino} and M. {Furdek} and S. {Popov} and O. {Ozolins}},
  journal={IEEE/OSA Journal of Optical Communications and Networking}, 
  title={Fast signal quality monitoring for coherent communications enabled by CNN-based EVM estimation}, 
  year={2021},
  volume={13},
  number={4},
  pages={B12-B20},
  doi={10.1364/JOCN.409704}
}

Dataset:

@data{FanEtAl:Dataset:EVM:2021,
    doi = {10.21227/1684-a275},
    url = {https://dx.doi.org/10.21227/1684-a275},
    author={Y. {Fan} and A. {Udalcovs} and X. {Pang} and C. {Natalino} and M. {Furdek} and S. {Popov} and O. {Ozolins}},
    publisher = {IEEE Dataport},
    title = {2020_JOCN_Constellation_Dataset},
    year = {2020}
}

Running this code

To be able to run the Python file in this repository, you should have a Python environment with version 3.7 (it might work with newer versions) and the following packages:

• TensorFlow 2.x (tested with TF 2.4.1)
• Matplotlib
• imageio
• Scikit-Learn

You should also download the dataset mentioned above and unzip it within this project folder.

Contact

This repository is a fork from Yuchuan Fan's repository and is maintained by Carlos Natalino [Twitter], who can be contacted through [email protected].