SHCP: A Shallow Brain-Inspired Hierarchical Network for Near-Field Channel Prediction in ELAA Systems
IEEE Transactions on Communications · 2026 · 74 · 15895-15908
DOI: 10.1109/tcomm.2026.3745130 ↗Kurz gefasst
Fachartikel in der Zeitschrift IEEE Transactions on Communications, 2026. Verfasst von Yandi Cao und 2 weiteren Autor:innen. Die Arbeit nennt 41 Literaturangaben. Gefördert von: Mobile Information Network-National Science and Technology Major Project, Major Program (JD) of Hubei Province, Fundamental Research Funds for the Central Universities and the National Natural Science Foundation of China.
Abstract (Originalsprache)
In extremely large antenna array (ELAA) systems, the enlarged array aperture gives rise to spherical-wave propagation and spatial non-stationarity, posing significant challenges to conventional channel prediction methods. To address this problem, we draw inspiration from the shallow brain hypothesis (SBH) in neuroscience and propose a shallow brain-inspired hierarchical channel prediction (SHCP) network for near-field channel prediction in ELAA systems. SHCP organizes channel processing into three functional modules corresponding to the thalamic, cortical, and subcortical stages. Specifically, the thalamic module enhances historical CSI representations and forwards them to the downstream modules. The cortical module employs multiple customized attention pathways with cross-pathway interactions to capture the complex dependencies in channel evolution. The subcortical module integrates the cortical outputs with the thalamic features to produce the final prediction. By replacing conventional deep stacking with shallow parallel processing, SHCP enhances the representational diversity of the network while reducing the inference depth. Extensive simulations under ELAA channel conditions demonstrate that the proposed SHCP outperforms existing model-driven and learning-based baselines in prediction accuracy. Meanwhile, SHCP achieves lower computational complexity in terms of FLOPs and inference latency. These results indicate that the shallow brain-inspired architecture provides an effective framework for near-field channel prediction in ELAA systems.
Förderung
- Mobile Information Network-National Science and Technology Major Project
- Major Program (JD) of Hubei Province
- Fundamental Research Funds for the Central Universities and the National Natural Science Foundation of China
Zitieren
APA
Cao, Y., Yin, H., & Li, W. (2026). SHCP: A Shallow Brain-Inspired Hierarchical Network for Near-Field Channel Prediction in ELAA Systems. IEEE Transactions on Communications, 74, 15895-15908. https://doi.org/10.1109/tcomm.2026.3745130
Vancouver
Cao Y, Yin H, Li W. SHCP: A Shallow Brain-Inspired Hierarchical Network for Near-Field Channel Prediction in ELAA Systems. IEEE Transactions on Communications. 2026;74:15895-15908. doi:10.1109/tcomm.2026.3745130
BibTeX
@article{Cao2026,
author = {Cao, Yandi and Yin, Haifan and Li, Weidong},
title = {SHCP: A Shallow Brain-Inspired Hierarchical Network for Near-Field Channel Prediction in ELAA Systems},
journal = {IEEE Transactions on Communications},
year = {2026},
volume = {74},
pages = {15895--15908},
doi = {10.1109/tcomm.2026.3745130}
}
EUVIMED