This repository hosts code accompanying the paper [1] and the full source code will be released upon completed reviews.
The code is based on PyTorch.
The implementation is based on [2].
- The algorithm is susceptible to channel estimation noise
- We have proposed a learning-based approach that has a higher tolerance to estimation noise in [1].
- The algorithm experiences numerical instability when the number of segments increases (N>10)
- Unzip the package to your local directory, then
- Run 'pip install -r requirements.txt' to download the required packages;
main(num_segments = 5)
If you decide to use the source code for your research, please make sure to cite the original paper and ours:
- [1] M. Farsi, C. Häger, M. Karlsson, E. Agrell, "[Learning to Extract Distributed Polarization Sensing Data from Noisy Jones Matrices]", in proc. Optical Fiber Communication Conference (submitted), 2024.
- [2] R. Noé et al, "Polarization-Dependent Loss: New Definition and Measurement Techniques", in Journal of Lightwave Technology, vol. 33, no. 10, pp. 2127-2138, 2015.