EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Learned-SBL-GAMP based hybrid precoders/combiners in millimeter wave massive MIMO systems

Shoukath Ali K iD, Arfat Ahmad Khan iD, Perarasi T, Ateeq Ur Rehman, Khmaies Ouahada iD

PLOS ONE · 2023

DOI: 10.1371/journal.pone.0289868 ↗

Kurz gefasst

Fachartikel in der Zeitschrift PLOS ONE, 2023. Verfasst von Shoukath Ali K und 4 weiteren Autor:innen. Beteiligte Einrichtung: Department of Electronics and Communication Engineering, Presidency University, Itgalpura, Rajanukunte, Yelahanka, Bengaluru, Karnataka, India. Hauptthemen laut MEDLINE: Learning, Algorithms. Die Arbeit wurde bisher 10-mal zitiert (Crossref) und nennt 37 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: University of Johannesburg.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

In Millimeter-Wave (mm-Wave) massive Multiple-Input Multiple-Output (MIMO) systems, hybrid precoders/combiners must be designed to improve antenna gain and reduce hardware complexity. Sparse Bayesian learning via Expectation Maximization (SBL-EM) algorithm is not practically feasible for high signal dimensions because estimating sparse signals and designing optimal hybrid precoders/combiners using SBL-EM still provide high computational complexity for higher signal dimensions. To overcome the issues of high computational complexity along with making it suitable for larger data sets, in this paper, we propose Learned-Sparse Bayesian Learning with Generalized Approximate Message Passing algorithm (L-SBL-GAMP) algorithm for designing optimal hybrid precoders/combiners suitable for mmWave Massive MIMO systems. The L-SBL-GAMP algorithm is an extension of the SBL-GAMP algorithm that incorporates a Deep Neural Network (DNN) to improve the system performance. Based on the nature of the training data, the L-SBL-GAMP can design the optimal Hybrid precoders/combiners, which enhances the spectral efficiency of mmWave massive MIMO systems. The proposed L-SBL-GAMP algorithm reduces the iterations, training overhead, and computational complexity compared to the SBL-EM algorithm. The simulation results unveil that the proposed L-SBL-GAMP provides higher achievable rates, better accuracy, and low computational complexity compared to the existing algorithm, such as Orthogonal Matching Pursuit (OMP), Simultaneous Orthogonal Matching Pursuit (SOMP), SBL-EM and SBL-GAMP for mmWave massive MIMO architectures.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

LearningAlgorithmsBayes TheoremComputer SimulationNeural Networks, Computer

Autor:innen und Institutionen

  • Department of Electronics and Communication Engineering, Presidency University, Itgalpura, Rajanukunte, Yelahanka, Bengaluru, Karnataka, India.
  • Department of Computer Science, College of Computing, Khon Kaen University, Khon Kaen, Thailand.
  • Department of Electronics and Communication Engineering, Bannari Amman Institute of Technology, Anna University, Erode, Tamil Nadu, India.
  • Department of Electrical Engineering, Government College University, Lahore, Pakistan.
  • Department of Electrical and Electronic Engineering Science, University of Johannesburg, Johannesburg, South Africa.

ORCID-Profile: Shoukath Ali K, Arfat Ahmad Khan, Khmaies Ouahada

Förderung

  • University of Johannesburg

Zitieren

APA

Ali K, S., Khan, A. A., T, P., Ur Rehman, A., & Ouahada, K. (2023). Learned-SBL-GAMP based hybrid precoders/combiners in millimeter wave massive MIMO systems. PLOS ONE. https://doi.org/10.1371/journal.pone.0289868

Vancouver

Ali K S, Khan AA, T P, Ur Rehman A, Ouahada K. Learned-SBL-GAMP based hybrid precoders/combiners in millimeter wave massive MIMO systems. PLOS ONE. 2023. doi:10.1371/journal.pone.0289868

BibTeX
@article{AliK2023,
  author = {Ali K, Shoukath and Khan, Arfat Ahmad and T, Perarasi and Ur Rehman, Ateeq and Ouahada, Khmaies},
  title = {Learned-SBL-GAMP based hybrid precoders/combiners in millimeter wave massive MIMO systems},
  journal = {PLOS ONE},
  year = {2023},
  doi = {10.1371/journal.pone.0289868}
}
Kostenlos starten →