Spectrum Anomaly Detection for Optical Network Monitoring using Deep Unsupervised Learning

Authors: Carlos Natalino*, Aleksejs Udalcovs**, Lena Wosinska*, Oskars Ozolins**, Marija Furdek*

* Department of Electrical Engineering, Chalmers University of Technology, Gothenburg, Sweden (e-mail: [email protected], [email protected], [email protected]).

** RISE Research Institutes of Sweden, Kista, Sweden (e-mail: [email protected], [email protected]).

Abstract: Accurate and efficient anomaly detection is a key enabler for the cognitive management of optical networks, but traditional anomaly detection algorithms are computationally complex and do not scale well with the amount of monitoring data. Therefore, we propose an optical spectrum anomaly detection scheme that exploits computer vision and deep unsupervised learning to perform optical network monitoring relying only on constellation diagrams of received signals. The proposed scheme achieves 100% detection accuracy even without prior knowledge of the anomalies. Furthermore, operation with encoded images of constellation diagrams reduces the runtime by up to 200 times.

Paper (IEEEXplore): https://ieeexplore.ieee.org/document/9336677

Dataset (IEEE Data Port): https://dx.doi.org/10.21227/g9s9-ba02

Authors' version of the paper: https://research.chalmers.se/en/publication/522246

What is in this repository?

To run this implementation, you should download the dataset from IEEE DataPort and point the base_folder variable to the location of the dataset.

This repository contains the following files:

• A notebook implementing the training of the autoencoder here
• A notebook implementing the training (for OCSVM) and performance evaluation for 16QAM@40dB, 64QAM@40dB and 16QAM@25dB
• A notebook implementing the runtime performance testing here
• A notebook plotting the results here

Citing this work:

@ARTICLE{NatalinoEtAl:2021:CommLetters,
  author={C. {Natalino} and A. {Udalcovs} and L. {Wosinska} and O. {Ozolins} and M. {Furdek}},
  journal={IEEE Communications Letters}, 
  title={Spectrum Anomaly Detection for Optical Network Monitoring using Deep Unsupervised Learning}, 
  year={2021},
  pages={1-1},
  doi={10.1109/LCOMM.2021.3055064}
}

Dependencies:

This implementation was executed using Python 3.7 and the following libraries:

• TensorFlow 2.x
• Matplotlib
• Scikit-Learn
• ImageIO