WGAN-based Autoencoder Training Over-the-air
Preprint
Abstract
The practical realization of end-to-end training of communication systems is fundamentally limited by its accessibility of the channel gradient. To overcome this major burden, the idea of generative adversarial networks (GANs) that learn to mimic the actual channel behavior has been recently proposed in the literature. Contrarily to handcrafted classical channel modeling, which can never fully capture the real world, GANs promise, in principle, the ability to learn any physical impairment, enabled by the data-driven learning algorithm. In this work, we verify the concept of GAN-based autoencoder training in actual over-the-air (OTA) measurements. To improve training stability, we first extend the concept to conditional Wasserstein GANs and embed it into a state-of-the-art autoencoder-architecture, including bitwise estimates and an outer channel code. Further, in the same framework, we compare the existing three different training approaches: model-based pre-training with receiver finetuning, reinforcement learning (RL) and GAN-based channel modeling. For this, we show advantages and limitations of GAN-based end-to-end training. In particular, for non-linear effects, it turns out that learning the whole exploration space becomes prohibitively complex. Finally, we show that the training strategy benefits from a simpler (training) data acquisition when compared to RL-based training, which requires continuous transmitter weight updates. This becomes an important practical bottleneck due to limited bandwidth and latency between transmitter and training algorithm that may even operate at physically different locations.Keywords
All Related Versions
This publication has 13 references indexed in Scilit:
- Trainable Communication Systems: Concepts and PrototypeIEEE Transactions on Communications, 2020
- Model-Free Training of End-to-End Communication SystemsIEEE Journal on Selected Areas in Communications, 2019
- Approximating the Void: Learning Stochastic Channel Models from Observation with Variational Generative Adversarial Networks2019
- Online Label Recovery for Deep Learning-based Communication through Error Correcting Codes2018
- OFDM-Autoencoder for End-to-End Learning of Communications Systems2018
- Deep Learning Based Communication Over the AirIEEE Journal of Selected Topics in Signal Processing, 2017
- An Introduction to Deep Learning for the Physical LayerIEEE Transactions on Cognitive Communications and Networking, 2017
- Deep Residual Learning for Image Recognition2016
- The Earth Mover's Distance as a Metric for Image RetrievalInternational Journal of Computer Vision, 2000
- An artificial neural net Viterbi decoderIEEE Transactions on Communications, 1996