Graph Neural Network-Based Hybrid Precoding for Intelligent Integrated mmWave MIMO Systems
IEEE Transactions on Consumer Electronics · 2026-05 · 72 (2) · 4431-4440
DOI: 10.1109/tce.2026.3680485 ↗Kurz gefasst
Fachartikel in der Zeitschrift IEEE Transactions on Consumer Electronics, 2026-05. Verfasst von Lalitha Nagapuri und 1 weiteren Autor:innen. Die Arbeit nennt 27 Literaturangaben.
Abstract (Originalsprache)
This paper investigates graph neural network (GNN)-based hybrid precoding in millimeter-wave (mmWave) multiple-input multiple-output (MIMO) systems. The design of hybrid beamforming is formulated as a problem of graph representation, where the channel’s spatial and structural features are modeled by graph connections between antennas and data streams. Further, a GNN framework is used to approximate the optimal fully digital precoder by exploiting message-passing mechanisms that learn channel correlations. The proposed design is compared against conventional orthogonal matching pursuit (OMP) and deep learning-based convolutional neural network (CNN) methods. Simulation results show that the GNN-based method attains superior spectral efficiency and lower bit error rates in the considered signal-to-noise ratio (SNR) regime. In the high-SNR regime, the proposed scheme improves spectral efficiency by up to $3.67\times $ and $11.45\times $ compared to CNN and OMP, respectively, while approaching the performance of fully digital precoding under practical hardware constraints. These results confirm that GNN can be a powerful tool for the optimization of hybrid beamforming in next-generation mmWave MIMO systems.
Zitieren
APA
Nagapuri, L., & Banavathu, N. R. (2026). Graph Neural Network-Based Hybrid Precoding for Intelligent Integrated mmWave MIMO Systems. IEEE Transactions on Consumer Electronics, 72(2), 4431-4440. https://doi.org/10.1109/tce.2026.3680485
Vancouver
Nagapuri L, Banavathu NR. Graph Neural Network-Based Hybrid Precoding for Intelligent Integrated mmWave MIMO Systems. IEEE Transactions on Consumer Electronics. 2026;72(2):4431-4440. doi:10.1109/tce.2026.3680485
BibTeX
@article{Nagapuri2026,
author = {Nagapuri, Lalitha and Banavathu, Narasimha Rao},
title = {Graph Neural Network-Based Hybrid Precoding for Intelligent Integrated mmWave MIMO Systems},
journal = {IEEE Transactions on Consumer Electronics},
year = {2026},
volume = {72},
number = {2},
pages = {4431--4440},
doi = {10.1109/tce.2026.3680485}
}
EUVIMED