Meta-Autoencoder without Channel Model

This repository contains code for "End-to-End Fast Training of Communication Links Without a Channel Model via Online Meta-Learning" - Sangwoo Park, Osvaldo Simeone, and Joonhyuk Kang.

Dependencies

This program is written in python 3.7 and uses PyTorch 1.2.0 and scipy. Tensorboard for pytorch is used for visualization (e.g., https://pytorch.org/docs/stable/tensorboard.html).

• pip install tensorboard and pip install scipy might be useful.

Basic Usage under Rayleigh Block Fading channel case

• In the 'run/' folder, all the schemes in the paper can be trained via correlated channel model and tested with the same channels as done in the paper.