Decentralized Beamforming for Cell-Free Massive MIMO with Unsupervised Learning

TL;DR

This paper introduces unsupervised deep neural network architectures for decentralized beamforming in cell-free massive MIMO systems, significantly reducing signaling and computational complexity while maintaining near-optimal spectral efficiency.

Contribution

It presents novel fully and partially distributed unsupervised DNN architectures for decentralized beamforming in CF-mMIMO, minimizing communication overhead and computational load.

Findings

Achieve near-optimal sum-rate performance.

Reduce computational complexity by 10-24x.

Operate with zero or limited communication between APs and NC.

Abstract

Cell-free massive MIMO (CF-mMIMO) systems represent a promising approach to increase the spectral efficiency of wireless communication systems. However, near-optimal beamforming solutions require a large amount of signaling exchange between access points (APs) and the network controller (NC). In this letter, we propose two unsupervised deep neural networks (DNN) architectures, fully and partially distributed, that can perform decentralized coordinated beamforming with zero or limited communication overhead between APs and NC, for both fully digital and hybrid precoding. The proposed DNNs achieve near-optimal sum-rate while also reducing computational complexity by 10-24x compared to conventional near-optimal solutions.