Beamforming Design Via GNN in mmWave Cell-Free Massive MIMO Using Sub-6 GHz CSI
2026-08-31 • Machine Learning
Machine Learning
AI summaryⓘ
The authors study a way to improve communication in advanced wireless systems that use very high-frequency signals (mmWave) but require a lot of information about the signal environment. They show that instead of directly using this hard-to-get high-frequency data, they can use easier-to-get lower-frequency data (sub-6 GHz) to teach a special kind of neural network called a graph neural network (GNN) how to optimize the signal beams. Their method models the network as a graph and uses message passing to understand how users and base stations affect each other. Simulations show their approach performs as well or better than traditional methods that need full high-frequency information.
Millimeter-wave (mmWave)Cell-free massive MIMOBeamformingChannel state information (CSI)Graph neural network (GNN)Sub-6 GHzDownlink sum-rateMessage passingInter-user interferenceBase station cooperation
Authors
Sina Tavakolian, Abolfazl Zakeri, Ahmed Alkhateeb, Markku Juntti, Nhan Thanh Nguyen
Abstract
Beamforming methods in millimeter-wave (mmWave) cell-free massive multiple-input multiple-output (CFmMIMO) systems require accurate channel state information (CSI), whose acquisition entails significant training overhead. This paper shows that fully digital cell-free mmWave beamforming can be effectively learned from sub-6 GHz CSI using a graph neural network (GNN). Specifically, we represent a CFmMIMO system as a wireless graph, and the GNN is trained to approximate beamformers that maximize the downlink sum-rate based on the available sub-6 GHz CSI. A message-passing mechanism is proposed to capture inter-user interference and inter-base-station cooperation across different network topologies. Simulation results demonstrate that the proposed sub-6 GHz-assisted GNN-based beamformer achieves competitive and often superior sum-rate performance compared to classical baselines that rely on full mmWave CSI.