Core Concepts
Efficiently predict mmWave beamspace from sub-6G channel estimates for superior spectrum efficiency.
Abstract
The article introduces a Sub-6G information Aided Multi-User Hybrid Beamforming (SA-MUHBF) framework to optimize beamforming in multi-user scenarios. By leveraging spatial congruence between sub-6GHz and mmWave channels, the framework efficiently predicts mmWave beamspace representation, selects analog beams using a graph neural network, and designs digital beamforming. Numerical results show superior spectrum efficiency over benchmarks.
Stats
Numerical results demonstrate superior spectrum efficiency.
The DeepMIMO dataset is used for network training.