A Fast 3-D MIMO Imaging Method Based on 3-D Generalized Approximate Message Passing Pattern-Coupled Sparse Bayesian Learning
IEEE Transactions on Aerospace and Electronic Systems · 2026 · 62 · 14657-14672
DOI: 10.1109/taes.2026.3706473 ↗Kurz gefasst
Fachartikel in der Zeitschrift IEEE Transactions on Aerospace and Electronic Systems, 2026. Verfasst von Yuyang Shao und 7 weiteren Autor:innen. Die Arbeit nennt 42 Literaturangaben. Gefördert von: National Natural Science Foundation of China, National Defense Basic Scientific Research Program of China, Fund for Foreign Scholars in University Research and Teaching Programs.
Abstract (Originalsprache)
In 3-D multi-input-multi-output imaging, sparse Bayesian learning has been introduced to address the image resolution constraint caused by the finite array aperture. However, in practical imaging scenarios, targets are often concentrated in the center region of the scene rather than being randomly distributed, a structural characteristic that conventional sparse Bayesian learning (CSBL) fails to exploit. Therefore, we propose a 3-D block-sparse signal reconstruction method based on pattern-coupled sparse Bayesian learning. The proposed approach employs an improved pattern-coupled hierarchical Gaussian model as the sparse prior for 3-D image reconstruction. Specifically, the sparsity of each grid is determined by both its own hyperparameters and those of its 26 neighboring grids, rather than solely by its own hyperparameters as in CSBL. The proposed prior model provides greater flexibility in capturing the 3-D block sparse structures. At the same time, the increased computational complexity introduced by matrix inversion also limits its application. To overcome this limitation, generalized approximate message passing is incorporated into sparse Bayesian learning to accelerate the algorithm execution. Both simulation and experimental results demonstrate that the proposed algorithm can more effectively exploit the inherent 3-D structure information of the scene and achieves superior image reconstruction performance.
Förderung
- National Natural Science Foundation of China
- National Defense Basic Scientific Research Program of China
- Fund for Foreign Scholars in University Research and Teaching Programs
- National Key RD Program of China
- Postdoctoral Innovation Talent Support Program
Zitieren
APA
Shao, Y., Ma, H., Liu, H., Su, T., Li, C., Zhu, H., Peng, W., & Pang, G. (2026). A Fast 3-D MIMO Imaging Method Based on 3-D Generalized Approximate Message Passing Pattern-Coupled Sparse Bayesian Learning. IEEE Transactions on Aerospace and Electronic Systems, 62, 14657-14672. https://doi.org/10.1109/taes.2026.3706473
Vancouver
Shao Y, Ma H, Liu H, Su T, Li C, Zhu H, et al. A Fast 3-D MIMO Imaging Method Based on 3-D Generalized Approximate Message Passing Pattern-Coupled Sparse Bayesian Learning. IEEE Transactions on Aerospace and Electronic Systems. 2026;62:14657-14672. doi:10.1109/taes.2026.3706473
BibTeX
@article{Shao2026,
author = {Shao, Yuyang and Ma, Hui and Liu, Hongwei and Su, Tao and Li, Caipin and Zhu, Haoyue and Peng, Wencan and Pang, Guanghui},
title = {A Fast 3-D MIMO Imaging Method Based on 3-D Generalized Approximate Message Passing Pattern-Coupled Sparse Bayesian Learning},
journal = {IEEE Transactions on Aerospace and Electronic Systems},
year = {2026},
volume = {62},
pages = {14657--14672},
doi = {10.1109/taes.2026.3706473}
}
EUVIMED