PowerNetMax: A DRL-GNN framework for IRS-Assisted IOT network optimization

Publikation: Beitrag in FachzeitschriftArtikelForschungPeer-Review

Autorschaft

  • Muhammad Farhan
  • Lei Wang
  • Nadir Shah
  • Gabriel Miro Muntean
  • Awais Bin Asif
  • Houbing Herbert Song

Externe Organisationen

  • Dalian University of Technology
  • COMSATS Institute of Information Technology
  • Dublin City University
  • University of Maryland Baltimore County
Forschungs-netzwerk anzeigen

Details

OriginalspracheEnglisch
Aufsatznummer111760
FachzeitschriftComputer networks
Jahrgang273
Frühes Online-Datum9 Okt. 2025
PublikationsstatusVeröffentlicht - Dez. 2025

Abstract

Intelligent Reflecting Surfaces (IRS) have recently emerged as a cutting-edge technology in 6G Internet of Things (IoT) communications, offering substantial connectivity enhancements, particularly in remote, high-mobility, or obstacle-prone environments. This paper proposes PowerNetMax, an innovative framework designed to improve overall network connectivity, reliability, and energy efficiency in IRS-assisted IoT communication systems. PowerNetMax leverages a comprehensive set of network parameters and integrates the strengths of Deep Reinforcement Learning (DRL) and Graph Neural Networks (GNN) to enable intelligent and adaptive optimization. Through extensive experimentation, PowerNetMax demonstrates up to 5–20 % higher received power, 50 % faster convergence, and 20 % higher throughput under mobility conditions compared to state-of-the-art GNN-based and heuristic solutions. Extensive simulation results confirm that PowerNetMax achieves superior adaptability and robustness, highlighting its effectiveness for future IRS-assisted IoT networks.

ASJC Scopus Sachgebiete

Ziele für nachhaltige Entwicklung

Zitieren

PowerNetMax: A DRL-GNN framework for IRS-Assisted IOT network optimization. / Farhan, Muhammad; Wang, Lei; Shah, Nadir et al.
in: Computer networks, Jahrgang 273, 111760, 12.2025.

Publikation: Beitrag in FachzeitschriftArtikelForschungPeer-Review

Farhan, M, Wang, L, Shah, N, Muntean, GM, Asif, AB & Song, HH 2025, 'PowerNetMax: A DRL-GNN framework for IRS-Assisted IOT network optimization', Computer networks, Jg. 273, 111760. https://doi.org/10.1016/j.comnet.2025.111760
Farhan, M., Wang, L., Shah, N., Muntean, G. M., Asif, A. B., & Song, H. H. (2025). PowerNetMax: A DRL-GNN framework for IRS-Assisted IOT network optimization. Computer networks, 273, Artikel 111760. https://doi.org/10.1016/j.comnet.2025.111760
Farhan M, Wang L, Shah N, Muntean GM, Asif AB, Song HH. PowerNetMax: A DRL-GNN framework for IRS-Assisted IOT network optimization. Computer networks. 2025 Dez;273:111760. Epub 2025 Okt 9. doi: 10.1016/j.comnet.2025.111760
Farhan, Muhammad ; Wang, Lei ; Shah, Nadir et al. / PowerNetMax : A DRL-GNN framework for IRS-Assisted IOT network optimization. in: Computer networks. 2025 ; Jahrgang 273.
Download
@article{a99d78a1e8a14908818bc3d675bedccd,
title = "PowerNetMax: A DRL-GNN framework for IRS-Assisted IOT network optimization",
abstract = "Intelligent Reflecting Surfaces (IRS) have recently emerged as a cutting-edge technology in 6G Internet of Things (IoT) communications, offering substantial connectivity enhancements, particularly in remote, high-mobility, or obstacle-prone environments. This paper proposes PowerNetMax, an innovative framework designed to improve overall network connectivity, reliability, and energy efficiency in IRS-assisted IoT communication systems. PowerNetMax leverages a comprehensive set of network parameters and integrates the strengths of Deep Reinforcement Learning (DRL) and Graph Neural Networks (GNN) to enable intelligent and adaptive optimization. Through extensive experimentation, PowerNetMax demonstrates up to 5–20 % higher received power, 50 % faster convergence, and 20 % higher throughput under mobility conditions compared to state-of-the-art GNN-based and heuristic solutions. Extensive simulation results confirm that PowerNetMax achieves superior adaptability and robustness, highlighting its effectiveness for future IRS-assisted IoT networks.",
keywords = "Deep reinforcement learning, Graph neural network, Intelligent reflecting surface, Non-terrestrial network, Unmanned aerial vehicles",
author = "Muhammad Farhan and Lei Wang and Nadir Shah and Muntean, {Gabriel Miro} and Asif, {Awais Bin} and Song, {Houbing Herbert}",
note = "Publisher Copyright: {\textcopyright} 2025 Elsevier B.V.",
year = "2025",
month = dec,
doi = "10.1016/j.comnet.2025.111760",
language = "English",
volume = "273",
journal = "Computer networks",
issn = "1389-1286",
publisher = "North-Holland Publ Co",

}

Download

TY - JOUR

T1 - PowerNetMax

T2 - A DRL-GNN framework for IRS-Assisted IOT network optimization

AU - Farhan, Muhammad

AU - Wang, Lei

AU - Shah, Nadir

AU - Muntean, Gabriel Miro

AU - Asif, Awais Bin

AU - Song, Houbing Herbert

N1 - Publisher Copyright: © 2025 Elsevier B.V.

PY - 2025/12

Y1 - 2025/12

N2 - Intelligent Reflecting Surfaces (IRS) have recently emerged as a cutting-edge technology in 6G Internet of Things (IoT) communications, offering substantial connectivity enhancements, particularly in remote, high-mobility, or obstacle-prone environments. This paper proposes PowerNetMax, an innovative framework designed to improve overall network connectivity, reliability, and energy efficiency in IRS-assisted IoT communication systems. PowerNetMax leverages a comprehensive set of network parameters and integrates the strengths of Deep Reinforcement Learning (DRL) and Graph Neural Networks (GNN) to enable intelligent and adaptive optimization. Through extensive experimentation, PowerNetMax demonstrates up to 5–20 % higher received power, 50 % faster convergence, and 20 % higher throughput under mobility conditions compared to state-of-the-art GNN-based and heuristic solutions. Extensive simulation results confirm that PowerNetMax achieves superior adaptability and robustness, highlighting its effectiveness for future IRS-assisted IoT networks.

AB - Intelligent Reflecting Surfaces (IRS) have recently emerged as a cutting-edge technology in 6G Internet of Things (IoT) communications, offering substantial connectivity enhancements, particularly in remote, high-mobility, or obstacle-prone environments. This paper proposes PowerNetMax, an innovative framework designed to improve overall network connectivity, reliability, and energy efficiency in IRS-assisted IoT communication systems. PowerNetMax leverages a comprehensive set of network parameters and integrates the strengths of Deep Reinforcement Learning (DRL) and Graph Neural Networks (GNN) to enable intelligent and adaptive optimization. Through extensive experimentation, PowerNetMax demonstrates up to 5–20 % higher received power, 50 % faster convergence, and 20 % higher throughput under mobility conditions compared to state-of-the-art GNN-based and heuristic solutions. Extensive simulation results confirm that PowerNetMax achieves superior adaptability and robustness, highlighting its effectiveness for future IRS-assisted IoT networks.

KW - Deep reinforcement learning

KW - Graph neural network

KW - Intelligent reflecting surface

KW - Non-terrestrial network

KW - Unmanned aerial vehicles

UR - http://www.scopus.com/inward/record.url?scp=105019064361&partnerID=8YFLogxK

U2 - 10.1016/j.comnet.2025.111760

DO - 10.1016/j.comnet.2025.111760

M3 - Article

AN - SCOPUS:105019064361

VL - 273

JO - Computer networks

JF - Computer networks

SN - 1389-1286

M1 - 111760

ER -