Deep Compressed Sensing for Terahertz Ultra-Massive MIMO Channel Estimation

Envisioned as a pivotal technology for sixth-generation (6G) and beyond, Terahertz (THz) band communications can potentially satisfy the increasing demand for ultra-high-speed wireless links. While ultra-massive multiple-input multiple-output (UM-MIMO) is promising in counteracting the exceptionally...

Full description

Saved in:
Bibliographic Details
Main Authors: Ganghui Lin, Mikail Erdem, Mohamed-Slim Alouini
Format: Article
Language:English
Published: IEEE 2025-01-01
Series:IEEE Open Journal of the Communications Society
Subjects:
Online Access:https://ieeexplore.ieee.org/document/10899780/
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849328080146923520
author Ganghui Lin
Mikail Erdem
Mohamed-Slim Alouini
author_facet Ganghui Lin
Mikail Erdem
Mohamed-Slim Alouini
author_sort Ganghui Lin
collection DOAJ
description Envisioned as a pivotal technology for sixth-generation (6G) and beyond, Terahertz (THz) band communications can potentially satisfy the increasing demand for ultra-high-speed wireless links. While ultra-massive multiple-input multiple-output (UM-MIMO) is promising in counteracting the exceptionally high path loss at THz frequency, the channel estimation (CE) of this extensive antenna system introduces significant challenges. In this paper, we propose a deep compressed sensing (DCS) framework based on generative neural networks for THz CE. The proposed estimator generates realistic THz channel samples to avoid complex channel modeling for THz UM-MIMO systems, especially in the near field. More importantly, the estimator is optimized for fast channel inference. Our results show significant superiority over the baseline generative adversarial network (GAN) estimator and traditional estimators. Compared to conventional estimators, our model achieves at least 8 dB lower normalized mean squared error (NMSE). Against GAN estimator, our model achieves around 3 dB lower NMSE at 0 dB SNR with one order of magnitude lower computation complexity. Moreover, our model achieves lower training overhead compared to GAN with empirically 4 times faster training convergence.
format Article
id doaj-art-a2eb8a13f6354cfe97f3ea7abc841bca
institution Kabale University
issn 2644-125X
language English
publishDate 2025-01-01
publisher IEEE
record_format Article
series IEEE Open Journal of the Communications Society
spelling doaj-art-a2eb8a13f6354cfe97f3ea7abc841bca2025-08-20T03:47:41ZengIEEEIEEE Open Journal of the Communications Society2644-125X2025-01-0161747176210.1109/OJCOMS.2025.354487110899780Deep Compressed Sensing for Terahertz Ultra-Massive MIMO Channel EstimationGanghui Lin0https://orcid.org/0000-0002-3436-3737Mikail Erdem1https://orcid.org/0000-0003-3501-4229Mohamed-Slim Alouini2https://orcid.org/0000-0003-4827-1793Division of Computer, Electrical and Mathematical Sciences and Engineering, King Abdullah University of Science and Technology, Thuwal, Saudi ArabiaDivision of Computer, Electrical and Mathematical Sciences and Engineering, King Abdullah University of Science and Technology, Thuwal, Saudi ArabiaDivision of Computer, Electrical and Mathematical Sciences and Engineering, King Abdullah University of Science and Technology, Thuwal, Saudi ArabiaEnvisioned as a pivotal technology for sixth-generation (6G) and beyond, Terahertz (THz) band communications can potentially satisfy the increasing demand for ultra-high-speed wireless links. While ultra-massive multiple-input multiple-output (UM-MIMO) is promising in counteracting the exceptionally high path loss at THz frequency, the channel estimation (CE) of this extensive antenna system introduces significant challenges. In this paper, we propose a deep compressed sensing (DCS) framework based on generative neural networks for THz CE. The proposed estimator generates realistic THz channel samples to avoid complex channel modeling for THz UM-MIMO systems, especially in the near field. More importantly, the estimator is optimized for fast channel inference. Our results show significant superiority over the baseline generative adversarial network (GAN) estimator and traditional estimators. Compared to conventional estimators, our model achieves at least 8 dB lower normalized mean squared error (NMSE). Against GAN estimator, our model achieves around 3 dB lower NMSE at 0 dB SNR with one order of magnitude lower computation complexity. Moreover, our model achieves lower training overhead compared to GAN with empirically 4 times faster training convergence.https://ieeexplore.ieee.org/document/10899780/Channel estimationgenerative neural networkTerahertzultra-massive MIMO
spellingShingle Ganghui Lin
Mikail Erdem
Mohamed-Slim Alouini
Deep Compressed Sensing for Terahertz Ultra-Massive MIMO Channel Estimation
IEEE Open Journal of the Communications Society
Channel estimation
generative neural network
Terahertz
ultra-massive MIMO
title Deep Compressed Sensing for Terahertz Ultra-Massive MIMO Channel Estimation
title_full Deep Compressed Sensing for Terahertz Ultra-Massive MIMO Channel Estimation
title_fullStr Deep Compressed Sensing for Terahertz Ultra-Massive MIMO Channel Estimation
title_full_unstemmed Deep Compressed Sensing for Terahertz Ultra-Massive MIMO Channel Estimation
title_short Deep Compressed Sensing for Terahertz Ultra-Massive MIMO Channel Estimation
title_sort deep compressed sensing for terahertz ultra massive mimo channel estimation
topic Channel estimation
generative neural network
Terahertz
ultra-massive MIMO
url https://ieeexplore.ieee.org/document/10899780/
work_keys_str_mv AT ganghuilin deepcompressedsensingforterahertzultramassivemimochannelestimation
AT mikailerdem deepcompressedsensingforterahertzultramassivemimochannelestimation
AT mohamedslimalouini deepcompressedsensingforterahertzultramassivemimochannelestimation