GNN-Enhanced Approximate Message Passing for Massive/Ultra-Massive MIMO Detection

Hengtao He; Alva Kosasih; Xianghao Yu; Jun Zhang; S. H. Song; Wibowo Hardjawana; Khaled B. Letaief

doi:10.1109/WCNC55385.2023.10118792

GNN-Enhanced Approximate Message Passing for Massive/Ultra-Massive MIMO Detection

Hengtao He, Alva Kosasih, Xianghao Yu, Jun Zhang, S. H. Song, Wibowo Hardjawana, Khaled B. Letaief

Department of Electrical Engineering

Research output: Chapters, Conference Papers, Creative and Literary Works › RGC 32 - Refereed conference paper (with host publication) › peer-review

12 Citations (Scopus)

Abstract

Efficient massive/ultra-massive multiple-input multiple-output (MIMO) detection algorithms with satisfactory performance and low complexity are critical to meet the high throughput and ultra-low latency requirements in 5G and beyond communications, given the extremely large number of antennas. In this paper, we propose a low complexity graph neural network (GNN) enhanced approximate message passing (AMP) algorithm, AMP-GNN, for massive/ultra-massive MIMO detection. The structure of the neural network is customized by unfolding the AMP algorithm and introducing the GNN module for multiuser interference cancellation. Numerical results will show that the proposed AMP-GNN significantly improves the performance of the AMP detector and achieves comparable performance as the state-of-the-art deep learning-based MIMO detectors but with reduced computational complexity. Furthermore, it presents strong robustness to the change of the number of users. © 2023 IEEE.

Original language	English
Title of host publication	2023 IEEE Wireless Communications and Networking Conference (WCNC)
Subtitle of host publication	Proceedings
Publisher	IEEE
Number of pages	6
ISBN (Electronic)	9781665491228
ISBN (Print)	978-1-6654-9123-5
DOIs	https://doi.org/10.1109/WCNC55385.2023.10118792
Publication status	Published - 2023
Event	2023 IEEE Wireless Communications and Networking Conference (WCNC 2023) - Glasgow, United Kingdom Duration: 26 Mar 2023 → 29 Mar 2023 https://wcnc2023.ieee-wcnc.org/

Publication series

Name	IEEE Wireless Communications and Networking Conference, WCNC
ISSN (Print)	1525-3511
ISSN (Electronic)	1558-2612

Conference

Conference	2023 IEEE Wireless Communications and Networking Conference (WCNC 2023)
Country/Territory	United Kingdom
City	Glasgow
Period	26/03/23 → 29/03/23
Internet address	https://wcnc2023.ieee-wcnc.org/

Funding

This work was supported by the General Research Fund (Projects No. 16212120, 16212922) and Research Impact Fund (Project No. R5009-21) from the Hong Kong Research Grants Council. The work was also supported by the Shenzhen Science and Technology Innovation Committee under Grant SGDX20210823103201006.

Access Output

10.1109/WCNC55385.2023.10118792

Other Access Options

Check CityUHK Library

Permanent Link

Right click to copy link

Cite this

He, H., Kosasih, A., Yu, X., Zhang, J., Song, S. H., Hardjawana, W., & Letaief, K. B. (2023). GNN-Enhanced Approximate Message Passing for Massive/Ultra-Massive MIMO Detection. In 2023 IEEE Wireless Communications and Networking Conference (WCNC): Proceedings (IEEE Wireless Communications and Networking Conference, WCNC). IEEE. https://doi.org/10.1109/WCNC55385.2023.10118792

@inproceedings{f29daf286eb940a5bca505ea3267e861,

title = "GNN-Enhanced Approximate Message Passing for Massive/Ultra-Massive MIMO Detection",

abstract = "Efficient massive/ultra-massive multiple-input multiple-output (MIMO) detection algorithms with satisfactory performance and low complexity are critical to meet the high throughput and ultra-low latency requirements in 5G and beyond communications, given the extremely large number of antennas. In this paper, we propose a low complexity graph neural network (GNN) enhanced approximate message passing (AMP) algorithm, AMP-GNN, for massive/ultra-massive MIMO detection. The structure of the neural network is customized by unfolding the AMP algorithm and introducing the GNN module for multiuser interference cancellation. Numerical results will show that the proposed AMP-GNN significantly improves the performance of the AMP detector and achieves comparable performance as the state-of-the-art deep learning-based MIMO detectors but with reduced computational complexity. Furthermore, it presents strong robustness to the change of the number of users. {\textcopyright} 2023 IEEE.",

author = "Hengtao He and Alva Kosasih and Xianghao Yu and Jun Zhang and Song, {S. H.} and Wibowo Hardjawana and Letaief, {Khaled B.}",

year = "2023",

doi = "10.1109/WCNC55385.2023.10118792",

language = "English",

isbn = "978-1-6654-9123-5",

series = "IEEE Wireless Communications and Networking Conference, WCNC",

publisher = "IEEE",

booktitle = "2023 IEEE Wireless Communications and Networking Conference (WCNC)",

address = "United States",

note = "2023 IEEE Wireless Communications and Networking Conference (WCNC 2023) ; Conference date: 26-03-2023 Through 29-03-2023",

url = "https://wcnc2023.ieee-wcnc.org/",

}

He, H, Kosasih, A, Yu, X, Zhang, J, Song, SH, Hardjawana, W & Letaief, KB 2023, GNN-Enhanced Approximate Message Passing for Massive/Ultra-Massive MIMO Detection. in 2023 IEEE Wireless Communications and Networking Conference (WCNC): Proceedings. IEEE Wireless Communications and Networking Conference, WCNC, IEEE, 2023 IEEE Wireless Communications and Networking Conference (WCNC 2023), Glasgow, United Kingdom, 26/03/23. https://doi.org/10.1109/WCNC55385.2023.10118792

GNN-Enhanced Approximate Message Passing for Massive/Ultra-Massive MIMO Detection. / He, Hengtao; Kosasih, Alva; Yu, Xianghao et al.
2023 IEEE Wireless Communications and Networking Conference (WCNC): Proceedings. IEEE, 2023. (IEEE Wireless Communications and Networking Conference, WCNC).

Research output: Chapters, Conference Papers, Creative and Literary Works › RGC 32 - Refereed conference paper (with host publication) › peer-review

TY - GEN

T1 - GNN-Enhanced Approximate Message Passing for Massive/Ultra-Massive MIMO Detection

AU - He, Hengtao

AU - Kosasih, Alva

AU - Yu, Xianghao

AU - Zhang, Jun

AU - Song, S. H.

AU - Hardjawana, Wibowo

AU - Letaief, Khaled B.

PY - 2023

Y1 - 2023

N2 - Efficient massive/ultra-massive multiple-input multiple-output (MIMO) detection algorithms with satisfactory performance and low complexity are critical to meet the high throughput and ultra-low latency requirements in 5G and beyond communications, given the extremely large number of antennas. In this paper, we propose a low complexity graph neural network (GNN) enhanced approximate message passing (AMP) algorithm, AMP-GNN, for massive/ultra-massive MIMO detection. The structure of the neural network is customized by unfolding the AMP algorithm and introducing the GNN module for multiuser interference cancellation. Numerical results will show that the proposed AMP-GNN significantly improves the performance of the AMP detector and achieves comparable performance as the state-of-the-art deep learning-based MIMO detectors but with reduced computational complexity. Furthermore, it presents strong robustness to the change of the number of users. © 2023 IEEE.

AB - Efficient massive/ultra-massive multiple-input multiple-output (MIMO) detection algorithms with satisfactory performance and low complexity are critical to meet the high throughput and ultra-low latency requirements in 5G and beyond communications, given the extremely large number of antennas. In this paper, we propose a low complexity graph neural network (GNN) enhanced approximate message passing (AMP) algorithm, AMP-GNN, for massive/ultra-massive MIMO detection. The structure of the neural network is customized by unfolding the AMP algorithm and introducing the GNN module for multiuser interference cancellation. Numerical results will show that the proposed AMP-GNN significantly improves the performance of the AMP detector and achieves comparable performance as the state-of-the-art deep learning-based MIMO detectors but with reduced computational complexity. Furthermore, it presents strong robustness to the change of the number of users. © 2023 IEEE.

UR - http://www.scopus.com/inward/record.url?scp=85148729504&partnerID=8YFLogxK

UR - https://www.scopus.com/record/pubmetrics.uri?eid=2-s2.0-85148729504&origin=recordpage

U2 - 10.1109/WCNC55385.2023.10118792

DO - 10.1109/WCNC55385.2023.10118792

M3 - RGC 32 - Refereed conference paper (with host publication)

SN - 978-1-6654-9123-5

T3 - IEEE Wireless Communications and Networking Conference, WCNC

BT - 2023 IEEE Wireless Communications and Networking Conference (WCNC)

PB - IEEE

T2 - 2023 IEEE Wireless Communications and Networking Conference (WCNC 2023)

Y2 - 26 March 2023 through 29 March 2023

ER -

GNN-Enhanced Approximate Message Passing for Massive/Ultra-Massive MIMO Detection

Abstract

Publication series

Conference

Funding

Access Output

Other Access Options

Permanent Link

Other Files and Links

Fingerprint

GRF: Collaborative Sensing and Communications for Perceptive Millimeter-Wave Wireless Networks

Cite this