Efficient power allocation using graph neural networks and deep algorithm unfolding

Arindam Chowdhury; Gunjan Verma; Chirag Rao; Ananthram Swami; Santiago Segarra

doi:10.1109/ICASSP39728.2021.9415106

Efficient power allocation using graph neural networks and deep algorithm unfolding

Arindam Chowdhury, Gunjan Verma, Chirag Rao, Ananthram Swami, Santiago Segarra

Computational Coupled Physics

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

18 Scopus citations

Abstract

We study the problem of optimal power allocation in a single-hop ad hoc wireless network. In solving this problem, we propose a hybrid neural architecture inspired by the algorithmic unfolding of the iterative weighted minimum mean squared error (WMMSE) method, that we denote as unfolded WMMSE (UWMMSE). The learnable weights within UWMMSE are parameterized using graph neural networks (GNNs), where the time-varying underlying graphs are given by the fading interference coefficients in the wireless network. These GNNs are trained through a gradient descent approach based on multiple instances of the power allocation problem. Once trained, UWMMSE achieves performance comparable to that of WMMSE while significantly reducing the computational complexity. This phenomenon is illustrated through numerical experiments along with the robustness and generalization to wireless networks of different densities and sizes.

Original language	English
Title of host publication	2021 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2021 - Proceedings
Publisher	Institute of Electrical and Electronics Engineers Inc.
Pages	4725-4729
Number of pages	5
ISBN (Electronic)	9781728176055
DOIs	https://doi.org/10.1109/ICASSP39728.2021.9415106
State	Published - 2021
Event	2021 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2021 - Virtual, Toronto, Canada Duration: Jun 6 2021 → Jun 11 2021

Publication series

Name	ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings
Volume	2021-June
ISSN (Print)	1520-6149

Conference

Conference	2021 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2021
Country/Territory	Canada
City	Virtual, Toronto
Period	06/6/21 → 06/11/21

Funding

Research was sponsored by the Army Research Office and was accomplished under Cooperative Agreement Number W911NF-19-2-0269. The views and conclusions contained in this document are those of the authors and should not be interpreted as representing the official policies, either expressed or implied, of the Army Research Office or the U.S. Government. The U.S. Government is authorized to reproduce and distribute reprints for Government purposes notwithstanding any copyright notation herein. Emails: {ac131, segarra}@rice.edu, {gunjan.verma.civ, chirag.r.rao.civ, ananthram.swami.civ}@mail.mil.

Keywords

Algorithm unfolding
Graph neural network
Power allocation
Wireless network
WMMSE

Access to Document

10.1109/ICASSP39728.2021.9415106

Cite this

Chowdhury, A., Verma, G., Rao, C., Swami, A., & Segarra, S. (2021). Efficient power allocation using graph neural networks and deep algorithm unfolding. In 2021 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2021 - Proceedings (pp. 4725-4729). (ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings; Vol. 2021-June). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/ICASSP39728.2021.9415106

Chowdhury, Arindam ; Verma, Gunjan ; Rao, Chirag et al. / Efficient power allocation using graph neural networks and deep algorithm unfolding. 2021 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2021 - Proceedings. Institute of Electrical and Electronics Engineers Inc., 2021. pp. 4725-4729 (ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings).

@inproceedings{c69e83f053ef44eb8b938db2b2534198,

title = "Efficient power allocation using graph neural networks and deep algorithm unfolding",

abstract = "We study the problem of optimal power allocation in a single-hop ad hoc wireless network. In solving this problem, we propose a hybrid neural architecture inspired by the algorithmic unfolding of the iterative weighted minimum mean squared error (WMMSE) method, that we denote as unfolded WMMSE (UWMMSE). The learnable weights within UWMMSE are parameterized using graph neural networks (GNNs), where the time-varying underlying graphs are given by the fading interference coefficients in the wireless network. These GNNs are trained through a gradient descent approach based on multiple instances of the power allocation problem. Once trained, UWMMSE achieves performance comparable to that of WMMSE while significantly reducing the computational complexity. This phenomenon is illustrated through numerical experiments along with the robustness and generalization to wireless networks of different densities and sizes.",

keywords = "Algorithm unfolding, Graph neural network, Power allocation, Wireless network, WMMSE",

author = "Arindam Chowdhury and Gunjan Verma and Chirag Rao and Ananthram Swami and Santiago Segarra",

note = "Publisher Copyright: {\textcopyright} 2021 IEEE.; 2021 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2021 ; Conference date: 06-06-2021 Through 11-06-2021",

year = "2021",

doi = "10.1109/ICASSP39728.2021.9415106",

language = "English",

series = "ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

pages = "4725--4729",

booktitle = "2021 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2021 - Proceedings",

}

Chowdhury, A, Verma, G, Rao, C, Swami, A & Segarra, S 2021, Efficient power allocation using graph neural networks and deep algorithm unfolding. in 2021 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2021 - Proceedings. ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings, vol. 2021-June, Institute of Electrical and Electronics Engineers Inc., pp. 4725-4729, 2021 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2021, Virtual, Toronto, Canada, 06/6/21. https://doi.org/10.1109/ICASSP39728.2021.9415106

Efficient power allocation using graph neural networks and deep algorithm unfolding. / Chowdhury, Arindam; Verma, Gunjan; Rao, Chirag et al.
2021 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2021 - Proceedings. Institute of Electrical and Electronics Engineers Inc., 2021. p. 4725-4729 (ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings; Vol. 2021-June).

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

TY - GEN

T1 - Efficient power allocation using graph neural networks and deep algorithm unfolding

AU - Chowdhury, Arindam

AU - Verma, Gunjan

AU - Rao, Chirag

AU - Swami, Ananthram

AU - Segarra, Santiago

PY - 2021

Y1 - 2021

N2 - We study the problem of optimal power allocation in a single-hop ad hoc wireless network. In solving this problem, we propose a hybrid neural architecture inspired by the algorithmic unfolding of the iterative weighted minimum mean squared error (WMMSE) method, that we denote as unfolded WMMSE (UWMMSE). The learnable weights within UWMMSE are parameterized using graph neural networks (GNNs), where the time-varying underlying graphs are given by the fading interference coefficients in the wireless network. These GNNs are trained through a gradient descent approach based on multiple instances of the power allocation problem. Once trained, UWMMSE achieves performance comparable to that of WMMSE while significantly reducing the computational complexity. This phenomenon is illustrated through numerical experiments along with the robustness and generalization to wireless networks of different densities and sizes.

AB - We study the problem of optimal power allocation in a single-hop ad hoc wireless network. In solving this problem, we propose a hybrid neural architecture inspired by the algorithmic unfolding of the iterative weighted minimum mean squared error (WMMSE) method, that we denote as unfolded WMMSE (UWMMSE). The learnable weights within UWMMSE are parameterized using graph neural networks (GNNs), where the time-varying underlying graphs are given by the fading interference coefficients in the wireless network. These GNNs are trained through a gradient descent approach based on multiple instances of the power allocation problem. Once trained, UWMMSE achieves performance comparable to that of WMMSE while significantly reducing the computational complexity. This phenomenon is illustrated through numerical experiments along with the robustness and generalization to wireless networks of different densities and sizes.

KW - Algorithm unfolding

KW - Graph neural network

KW - Power allocation

KW - Wireless network

KW - WMMSE

UR - http://www.scopus.com/inward/record.url?scp=85108567900&partnerID=8YFLogxK

U2 - 10.1109/ICASSP39728.2021.9415106

DO - 10.1109/ICASSP39728.2021.9415106

M3 - Conference contribution

AN - SCOPUS:85108567900

T3 - ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings

SP - 4725

EP - 4729

BT - 2021 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2021 - Proceedings

PB - Institute of Electrical and Electronics Engineers Inc.

T2 - 2021 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2021

Y2 - 6 June 2021 through 11 June 2021

ER -

Chowdhury A, Verma G, Rao C, Swami A, Segarra S. Efficient power allocation using graph neural networks and deep algorithm unfolding. In 2021 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2021 - Proceedings. Institute of Electrical and Electronics Engineers Inc. 2021. p. 4725-4729. (ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings). doi: 10.1109/ICASSP39728.2021.9415106

Efficient power allocation using graph neural networks and deep algorithm unfolding

Abstract

Publication series

Conference

Funding

Keywords

Access to Document

Other files and links

Fingerprint

Cite this