DM-DETECT – A Deep MIMO Detector for Beyond 5G Networks

Muhammad Yunis Daha; Joseph Rafferty; Muhammad Ikram Ashraf; Muhammad Usman Hadi

doi:10.1109/ICAISS58487.2023.10250495

DM-DETECT – A Deep MIMO Detector for Beyond 5G Networks

Muhammad Yunis Daha, Joseph Rafferty, Muhammad Ikram Ashraf, Muhammad Usman Hadi

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

3 Citations (Scopus)

Abstract

The revolution of Artificial Intelligence (AI) transforms the Multiple Input Multiple Output (MIMO) technology into Massive MIMO (Ma-MIMO) technology. However, despite the promising benefits of Ma-MIMO technology, it is very difficult to design a reliable and energy-efficient detector at the receiver end. To overcome this research challenge, this paper presents a new deep Ma-MIMO Detection (DM-DETECT) scheme for Ma-MIMO detection. The DM-DETECT uses deep learning (DL) to construct an AI-based network model. The DM-DETECT model is extensively trained and optimized to achieve high performance in realistic MIMO and Ma-MIMO use cases. Simulation results depict that the DM-DETECT performs better at a certain limit than the conventional detectors in terms of Symbol Error Rate (SER). Moreover, this study presents an in-depth analysis of the activation functions for deep neural networks at different SNR ranges, which offers valuable insights into optimizing the performance of the proposed DM-DETECT for Ma-MIMO technology.

Original language	English
Title of host publication	2023 Second International Conference on Augmented Intelligence and Sustainable Systems (ICAISS)
Publisher	IEEE Xplore
Pages	1381-1385
Number of pages	5
ISBN (Electronic)	979-8-3503-2579-9
ISBN (Print)	979-8-3503-2580-5
DOIs	https://doi.org/10.1109/ICAISS58487.2023.10250495
Publication status	Published online - 22 Sept 2023

Publication series

Name	Proceedings of the 2023 2nd International Conference on Augmented Intelligence and Sustainable Systems, ICAISS 2023

Bibliographical note

Funding Information:
ACKNOWLEDGMENT Muhammad Yunis Daha is supported by the Department of Economy (DfE) International at the School of Engineering, Ulster University, Belfast, United Kingdom.

Publisher Copyright:
© 2023 IEEE.

Keywords

5G and beyond
Multiple Input Multiple Output (MIMO)
MIMO Detection
Machine Learning
Deep Learning
symbol error rate
signal noise ratio

UN SDGs

This output contributes to the following UN Sustainable Development Goals (SDGs)

Access to Document

10.1109/ICAISS58487.2023.10250495

Cite this

Daha, M. Y., Rafferty, J., Ashraf, M. I., & Hadi, M. U. (2023). DM-DETECT – A Deep MIMO Detector for Beyond 5G Networks. In 2023 Second International Conference on Augmented Intelligence and Sustainable Systems (ICAISS) (pp. 1381-1385). (Proceedings of the 2023 2nd International Conference on Augmented Intelligence and Sustainable Systems, ICAISS 2023). IEEE Xplore. Advance online publication. https://doi.org/10.1109/ICAISS58487.2023.10250495

Daha, Muhammad Yunis ; Rafferty, Joseph ; Ashraf, Muhammad Ikram et al. / DM-DETECT – A Deep MIMO Detector for Beyond 5G Networks. 2023 Second International Conference on Augmented Intelligence and Sustainable Systems (ICAISS). IEEE Xplore, 2023. pp. 1381-1385 (Proceedings of the 2023 2nd International Conference on Augmented Intelligence and Sustainable Systems, ICAISS 2023).

@inproceedings{0318b48f86e647ee92e33cd294062bb3,

title = "DM-DETECT – A Deep MIMO Detector for Beyond 5G Networks",

abstract = "The revolution of Artificial Intelligence (AI) transforms the Multiple Input Multiple Output (MIMO) technology into Massive MIMO (Ma-MIMO) technology. However, despite the promising benefits of Ma-MIMO technology, it is very difficult to design a reliable and energy-efficient detector at the receiver end. To overcome this research challenge, this paper presents a new deep Ma-MIMO Detection (DM-DETECT) scheme for Ma-MIMO detection. The DM-DETECT uses deep learning (DL) to construct an AI-based network model. The DM-DETECT model is extensively trained and optimized to achieve high performance in realistic MIMO and Ma-MIMO use cases. Simulation results depict that the DM-DETECT performs better at a certain limit than the conventional detectors in terms of Symbol Error Rate (SER). Moreover, this study presents an in-depth analysis of the activation functions for deep neural networks at different SNR ranges, which offers valuable insights into optimizing the performance of the proposed DM-DETECT for Ma-MIMO technology.",

keywords = "5G and beyond, Multiple Input Multiple Output (MIMO), MIMO Detection, Machine Learning, Deep Learning, symbol error rate, signal noise ratio",

author = "Daha, {Muhammad Yunis} and Joseph Rafferty and Ashraf, {Muhammad Ikram} and Hadi, {Muhammad Usman}",

note = "Funding Information: ACKNOWLEDGMENT Muhammad Yunis Daha is supported by the Department of Economy (DfE) International at the School of Engineering, Ulster University, Belfast, United Kingdom. Publisher Copyright: {\textcopyright} 2023 IEEE.",

year = "2023",

month = sep,

day = "22",

doi = "10.1109/ICAISS58487.2023.10250495",

language = "English",

isbn = "979-8-3503-2580-5",

series = "Proceedings of the 2023 2nd International Conference on Augmented Intelligence and Sustainable Systems, ICAISS 2023",

publisher = "IEEE Xplore",

pages = "1381--1385",

booktitle = "2023 Second International Conference on Augmented Intelligence and Sustainable Systems (ICAISS)",

address = "United States",

}

Daha, MY, Rafferty, J, Ashraf, MI & Hadi, MU 2023, DM-DETECT – A Deep MIMO Detector for Beyond 5G Networks. in 2023 Second International Conference on Augmented Intelligence and Sustainable Systems (ICAISS). Proceedings of the 2023 2nd International Conference on Augmented Intelligence and Sustainable Systems, ICAISS 2023, IEEE Xplore, pp. 1381-1385. https://doi.org/10.1109/ICAISS58487.2023.10250495

DM-DETECT – A Deep MIMO Detector for Beyond 5G Networks. / Daha, Muhammad Yunis; Rafferty, Joseph; Ashraf, Muhammad Ikram et al.
2023 Second International Conference on Augmented Intelligence and Sustainable Systems (ICAISS). IEEE Xplore, 2023. p. 1381-1385 (Proceedings of the 2023 2nd International Conference on Augmented Intelligence and Sustainable Systems, ICAISS 2023).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - DM-DETECT – A Deep MIMO Detector for Beyond 5G Networks

AU - Daha, Muhammad Yunis

AU - Rafferty, Joseph

AU - Ashraf, Muhammad Ikram

AU - Hadi, Muhammad Usman

N1 - Funding Information: ACKNOWLEDGMENT Muhammad Yunis Daha is supported by the Department of Economy (DfE) International at the School of Engineering, Ulster University, Belfast, United Kingdom. Publisher Copyright: © 2023 IEEE.

PY - 2023/9/22

Y1 - 2023/9/22

N2 - The revolution of Artificial Intelligence (AI) transforms the Multiple Input Multiple Output (MIMO) technology into Massive MIMO (Ma-MIMO) technology. However, despite the promising benefits of Ma-MIMO technology, it is very difficult to design a reliable and energy-efficient detector at the receiver end. To overcome this research challenge, this paper presents a new deep Ma-MIMO Detection (DM-DETECT) scheme for Ma-MIMO detection. The DM-DETECT uses deep learning (DL) to construct an AI-based network model. The DM-DETECT model is extensively trained and optimized to achieve high performance in realistic MIMO and Ma-MIMO use cases. Simulation results depict that the DM-DETECT performs better at a certain limit than the conventional detectors in terms of Symbol Error Rate (SER). Moreover, this study presents an in-depth analysis of the activation functions for deep neural networks at different SNR ranges, which offers valuable insights into optimizing the performance of the proposed DM-DETECT for Ma-MIMO technology.

AB - The revolution of Artificial Intelligence (AI) transforms the Multiple Input Multiple Output (MIMO) technology into Massive MIMO (Ma-MIMO) technology. However, despite the promising benefits of Ma-MIMO technology, it is very difficult to design a reliable and energy-efficient detector at the receiver end. To overcome this research challenge, this paper presents a new deep Ma-MIMO Detection (DM-DETECT) scheme for Ma-MIMO detection. The DM-DETECT uses deep learning (DL) to construct an AI-based network model. The DM-DETECT model is extensively trained and optimized to achieve high performance in realistic MIMO and Ma-MIMO use cases. Simulation results depict that the DM-DETECT performs better at a certain limit than the conventional detectors in terms of Symbol Error Rate (SER). Moreover, this study presents an in-depth analysis of the activation functions for deep neural networks at different SNR ranges, which offers valuable insights into optimizing the performance of the proposed DM-DETECT for Ma-MIMO technology.

KW - 5G and beyond

KW - Multiple Input Multiple Output (MIMO)

KW - MIMO Detection

KW - Machine Learning

KW - Deep Learning

KW - symbol error rate

KW - signal noise ratio

UR - https://ieeexplore.ieee.org/xpl/conhome/1848470/all-proceedings

UR - https://ieeexplore.ieee.org/document/10250495

UR - http://www.scopus.com/inward/record.url?scp=85173611232&partnerID=8YFLogxK

U2 - 10.1109/ICAISS58487.2023.10250495

DO - 10.1109/ICAISS58487.2023.10250495

M3 - Conference contribution

SN - 979-8-3503-2580-5

T3 - Proceedings of the 2023 2nd International Conference on Augmented Intelligence and Sustainable Systems, ICAISS 2023

SP - 1381

EP - 1385

BT - 2023 Second International Conference on Augmented Intelligence and Sustainable Systems (ICAISS)

PB - IEEE Xplore

ER -

Daha MY, Rafferty J, Ashraf MI, Hadi MU. DM-DETECT – A Deep MIMO Detector for Beyond 5G Networks. In 2023 Second International Conference on Augmented Intelligence and Sustainable Systems (ICAISS). IEEE Xplore. 2023. p. 1381-1385. (Proceedings of the 2023 2nd International Conference on Augmented Intelligence and Sustainable Systems, ICAISS 2023). Epub 2023 Sept 22. doi: 10.1109/ICAISS58487.2023.10250495

DM-DETECT – A Deep MIMO Detector for Beyond 5G Networks

Abstract

Publication series

Bibliographical note

Keywords

UN SDGs

Access to Document

Other files and links

Fingerprint

Cite this