DeepTx: Deep Learning Beamforming with Channel Prediction

TL;DR

DeepTx introduces a deep learning-based beamforming method that predicts downlink channels from uplink estimates, enhancing wireless communication performance by learning channel evolution and compensating for errors.

Contribution

It presents a novel CNN model for transmitter-side beamforming that predicts downlink channels from uplink data, improving performance over traditional methods.

Findings

Enhanced beamforming performance demonstrated in numerical experiments.

CNN effectively predicts channel evolution and handles errors.

Supervised training based on UE receiver performance improves accuracy.

Abstract

Machine learning algorithms have recently been considered for many tasks in the field of wireless communications. Previously, we have proposed the use of a deep fully convolutional neural network (CNN) for receiver processing and shown it to provide considerable performance gains. In this study, we focus on machine learning algorithms for the transmitter. In particular, we consider beamforming and propose a CNN which, for a given uplink channel estimate as input, outputs downlink channel information to be used for beamforming. The CNN is trained in a supervised manner considering both uplink and downlink transmissions with a loss function that is based on UE receiver performance. The main task of the neural network is to predict the channel evolution between uplink and downlink slots, but it can also learn to handle inefficiencies and errors in the whole chain, including the actual beamforming phase. The provided numerical experiments demonstrate the improved beamforming performance.