Intensity Modulation/Direct Detection applying End-To-End Deep Learning Training

This repository implements a framework that enable us to train in an end-to-end fashion IM/DD system using PyTorch.

How to use it

poetry init
poetry install

You can run the simplest example included in exp_plain.py

poetry run python ./endtoend_plain/exps/exp_plain/exp_plain.py

The train routine requires a dictionary (named experimental dictionary) among with a models dictionary (including transmitter, channel and receiver). The get_exp_parameters.py demonstrates the basic construction of the experimental dictionary.

References (implementation applied)

• Kirtas, M., Passalis, N., Mourgias-Alexandris, G., Dabos, G., Pleros, N. and Tefas, A., 2022, August. Learning photonic neural network initialization for noise-aware end-to-end fiber transmission. In 2022 30th European signal processing conference (EUSIPCO) (pp. 1731-1735). IEEE.
• Kirtas, M., Passalis, N., Mourgias-Alexandris, G., Dabos, G., Pleros, N. and Tefas, A., 2022. Robust architecture-agnostic and noise resilient training of photonic deep learning models. IEEE Transactions on Emerging Topics in Computational Intelligence, 7(1), pp.140-149.
• Roumpos, I., Marinis, L.D., Kirtas, M., Passalis, N., Tefas, A., Contestabile, G., Pleros, N., Moralis-Pegios, M. and Vyrsokinos, K., 2023. High-performance end-to-end deep learning IM/DD link using optics-informed neural networks. Optics Express, 31(12), pp.20068-20079.
• De Marinis, L., Roumpos, I., Mourgias-Alexandris, G., Kirtas, M., Passalis, N., Tefas, A., Contestabile, G., Vyrsokinos, K., Pleros, N. and Moralis-Pegios, M., 2022, November. Improving noise resilience in end-to-end deep learning optical fiber transmission links. In 2022 Asia Communications and Photonics Conference (ACP) (pp. 1881-1884). IEEE.
• Roumpos, I., De Marinis, L., Mourgias-Alexandris, G., Kirtas, M., Passalis, N., Tefas, A., Contestabile, G., Vyrsokinos, K., Pleros, N. and Moralis-Pegios, M., 2023, May. Physics-inspired End-to-End Deep Learning for High-Performance Optical Fiber Transmission Links. In CLEO: Science and Innovations (pp. SF1F-6). Optica Publishing Group.