Graph Neural Network Based Access Point Selection for Cell-Free Massive MIMO Systems

TL;DR

This paper introduces a GNN-based AP selection algorithm for cell-free massive MIMO systems that improves prediction accuracy and scalability without relying on large-scale fading measurements.

Contribution

It develops a novel GNN approach using homogeneous and heterogeneous graphs for AP selection in cell-free MIMO, enhancing accuracy and scalability.

Findings

Outperforms proximity-based AP selection algorithms in accuracy.

Does not require measured signal strengths of all neighboring APs.

Scalable with the number of users in the system.

Abstract

A graph neural network (GNN) based access point (AP) selection algorithm for cell-free massive multiple-input multiple-output (MIMO) systems is proposed. Two graphs, a homogeneous graph which includes only AP nodes representing the structure of the APs in the network, and a heterogeneous graph which includes both AP nodes and user equipment (UE) nodes are constructed to represent a cell-free massive MIMO network. A GNN based on the inductive graph learning framework GraphSAGE is used to obtain the embeddings which are then used to predict the links between the nodes. The numerical results show that compared to the proximity-based AP selection algorithms, the proposed GNN based algorithm predicts the potential APs with more accuracy. Compared to the large scale fading coefficient based AP selection algorithms, the proposed algorithm does not require measured and sorted signal strengths of all the neighbouring APs. Furthermore, the proposed algorithm is scalable in terms of the number of users in the cell-free system.