Learning to Unfold Fractional Programming for Multi-Cell MU-MIMO Beamforming with Graph Neural Networks

TL;DR

This paper introduces a novel graph neural network-based approach to efficiently optimize beamforming in multi-cell MU-MIMO systems by unfolding the fractional programming algorithm, reducing computational complexity and improving convergence.

Contribution

It proposes a learning-based method that unfolds the FastFP algorithm using graph neural networks, offering a more efficient solution for beamforming optimization.

Findings

Reduces computational complexity compared to traditional FP methods.

Achieves faster convergence in beamforming optimization.

Demonstrates effectiveness through simulations on multi-cell MU-MIMO scenarios.

Abstract

In the multi-cell multiuser multi-input multi-output (MU-MIMO) systems, fractional programming (FP) has demonstrated considerable effectiveness in optimizing beamforming vectors, yet it suffers from high computational complexity. Recent improvements demonstrate reduced complexity by avoiding large-dimension matrix inversions (i.e., FastFP) and faster convergence by learning to unfold the FastFP algorithm (i.e., DeepFP).