Bit-wise Autoencoder Communication System Training
This video shows the training process of a bit-wise autoencoder communications system. The system tries to find optimal signals to transmit k=4 bit within n=2 real valued channel uses at the trained SNR of 3.85dB. On the left side one can see the currently used constellation and on the right side one can see the mutual information obtained at receiver side. The bottom heat map plots show the current decision regions of the learned receiver, where pink regions indicate a strong decision towards bit 0 and blue regions indicate a strong decision towards bit 1. It can be seen, that the system finally converges towards a constellation that differs from classical modulation schemes, like the comparable 16-QAM constellation, and is thereby able to transmit more information at this operational point. While these gains are well known in communications research as so called shaping gains, the possible game changer of a neural network based autoencoder system is, that it is able to optimize its signals over every kind of channel, including all unknown insufficiencies using data based training strategies.
For more information see the paper "Trainable Communication Systems: Concepts and Prototype" by S. Cammerer, F. A. Aoudia, S. Dörner, M. Stark, J. Hoydis and S. ten Brink at https://ieeexplore.ieee.org/document/... where these results are based on.
Citation:
S. Cammerer, F. A. Aoudia, S. Dörner, M. Stark, J. Hoydis and S. ten Brink, "Trainable Communication Systems: Concepts and Prototype," in IEEE Transactions on Communications, vol. 68, no. 9, pp. 5489-5503, Sept. 2020, doi: 10.1109/TCOMM.2020.3002915.
Institute of Telecommunications, University of Stuttgart