Efficient 5G Resource Block Scheduling Using Action Branching and Transformer Networks

Sylvain Nérondat; Xavier Leturc; Philippe Ciblat; Christophe J. Le Martret

doi:10.1109/ICMLCN64995.2025.11140453

Efficient 5G Resource Block Scheduling Using Action Branching and Transformer Networks

Sylvain Nérondat, Xavier Leturc, Philippe Ciblat, Christophe J. Le Martret

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

Abstract

This paper presents a deep reinforcement learning-based scheduling solution tailored for 5G networks. The proposed neural network architecture, utilizing an encoder-only transformer and action branching, is designed to handle large action spaces for resource block allocation in wireless environments. By training on variable number of user equipment scenarios, the solution generalizes well across different configurations. Experimental results in Nokia's wireless suite environment demonstrate superior performance in packet loss, compared to heuristics.

Original language	English
Title of host publication	2025 IEEE International Conference on Machine Learning for Communication and Networking, ICMLCN 2025
Publisher	Institute of Electrical and Electronics Engineers Inc.
ISBN (Electronic)	9798331520427
DOIs	https://doi.org/10.1109/ICMLCN64995.2025.11140453
Publication status	Published - 1 Jan 2025
Event	2nd IEEE International Conference on Machine Learning for Communication and Networking, ICMLCN 2025 - Barcelona, Spain Duration: 26 May 2025 → 29 May 2025

Publication series

Name	2025 IEEE International Conference on Machine Learning for Communication and Networking, ICMLCN 2025

Conference

Conference	2nd IEEE International Conference on Machine Learning for Communication and Networking, ICMLCN 2025
Country/Territory	Spain
City	Barcelona
Period	26/05/25 → 29/05/25

Keywords

Action branching
deep reinforcement learning
scheduling
transformer
wireless suite

Access to Document

10.1109/ICMLCN64995.2025.11140453

Cite this

Nérondat, S., Leturc, X., Ciblat, P., & Le Martret, C. J. (2025). Efficient 5G Resource Block Scheduling Using Action Branching and Transformer Networks. In 2025 IEEE International Conference on Machine Learning for Communication and Networking, ICMLCN 2025 (2025 IEEE International Conference on Machine Learning for Communication and Networking, ICMLCN 2025). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/ICMLCN64995.2025.11140453

Nérondat, Sylvain ; Leturc, Xavier ; Ciblat, Philippe et al. / Efficient 5G Resource Block Scheduling Using Action Branching and Transformer Networks. 2025 IEEE International Conference on Machine Learning for Communication and Networking, ICMLCN 2025. Institute of Electrical and Electronics Engineers Inc., 2025. (2025 IEEE International Conference on Machine Learning for Communication and Networking, ICMLCN 2025).

@inproceedings{8ade52b0e32c46688dc262642b6717d2,

title = "Efficient 5G Resource Block Scheduling Using Action Branching and Transformer Networks",

abstract = "This paper presents a deep reinforcement learning-based scheduling solution tailored for 5G networks. The proposed neural network architecture, utilizing an encoder-only transformer and action branching, is designed to handle large action spaces for resource block allocation in wireless environments. By training on variable number of user equipment scenarios, the solution generalizes well across different configurations. Experimental results in Nokia's wireless suite environment demonstrate superior performance in packet loss, compared to heuristics.",

keywords = "Action branching, deep reinforcement learning, scheduling, transformer, wireless suite",

author = "Sylvain N{\'e}rondat and Xavier Leturc and Philippe Ciblat and \{Le Martret\}, \{Christophe J.\}",

note = "Publisher Copyright: {\textcopyright} 2025 IEEE.; 2nd IEEE International Conference on Machine Learning for Communication and Networking, ICMLCN 2025 ; Conference date: 26-05-2025 Through 29-05-2025",

year = "2025",

month = jan,

day = "1",

doi = "10.1109/ICMLCN64995.2025.11140453",

language = "English",

series = "2025 IEEE International Conference on Machine Learning for Communication and Networking, ICMLCN 2025",

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

booktitle = "2025 IEEE International Conference on Machine Learning for Communication and Networking, ICMLCN 2025",

}

Nérondat, S, Leturc, X, Ciblat, P & Le Martret, CJ 2025, Efficient 5G Resource Block Scheduling Using Action Branching and Transformer Networks. in 2025 IEEE International Conference on Machine Learning for Communication and Networking, ICMLCN 2025. 2025 IEEE International Conference on Machine Learning for Communication and Networking, ICMLCN 2025, Institute of Electrical and Electronics Engineers Inc., 2nd IEEE International Conference on Machine Learning for Communication and Networking, ICMLCN 2025, Barcelona, Spain, 26/05/25. https://doi.org/10.1109/ICMLCN64995.2025.11140453

Efficient 5G Resource Block Scheduling Using Action Branching and Transformer Networks. / Nérondat, Sylvain; Leturc, Xavier; Ciblat, Philippe et al.
2025 IEEE International Conference on Machine Learning for Communication and Networking, ICMLCN 2025. Institute of Electrical and Electronics Engineers Inc., 2025. (2025 IEEE International Conference on Machine Learning for Communication and Networking, ICMLCN 2025).

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

TY - GEN

T1 - Efficient 5G Resource Block Scheduling Using Action Branching and Transformer Networks

AU - Nérondat, Sylvain

AU - Leturc, Xavier

AU - Ciblat, Philippe

AU - Le Martret, Christophe J.

PY - 2025/1/1

Y1 - 2025/1/1

N2 - This paper presents a deep reinforcement learning-based scheduling solution tailored for 5G networks. The proposed neural network architecture, utilizing an encoder-only transformer and action branching, is designed to handle large action spaces for resource block allocation in wireless environments. By training on variable number of user equipment scenarios, the solution generalizes well across different configurations. Experimental results in Nokia's wireless suite environment demonstrate superior performance in packet loss, compared to heuristics.

AB - This paper presents a deep reinforcement learning-based scheduling solution tailored for 5G networks. The proposed neural network architecture, utilizing an encoder-only transformer and action branching, is designed to handle large action spaces for resource block allocation in wireless environments. By training on variable number of user equipment scenarios, the solution generalizes well across different configurations. Experimental results in Nokia's wireless suite environment demonstrate superior performance in packet loss, compared to heuristics.

KW - Action branching

KW - deep reinforcement learning

KW - scheduling

KW - transformer

KW - wireless suite

UR - https://www.scopus.com/pages/publications/105016786148

U2 - 10.1109/ICMLCN64995.2025.11140453

DO - 10.1109/ICMLCN64995.2025.11140453

M3 - Conference contribution

AN - SCOPUS:105016786148

T3 - 2025 IEEE International Conference on Machine Learning for Communication and Networking, ICMLCN 2025

BT - 2025 IEEE International Conference on Machine Learning for Communication and Networking, ICMLCN 2025

PB - Institute of Electrical and Electronics Engineers Inc.

T2 - 2nd IEEE International Conference on Machine Learning for Communication and Networking, ICMLCN 2025

Y2 - 26 May 2025 through 29 May 2025

ER -

Nérondat S, Leturc X, Ciblat P, Le Martret CJ. Efficient 5G Resource Block Scheduling Using Action Branching and Transformer Networks. In 2025 IEEE International Conference on Machine Learning for Communication and Networking, ICMLCN 2025. Institute of Electrical and Electronics Engineers Inc. 2025. (2025 IEEE International Conference on Machine Learning for Communication and Networking, ICMLCN 2025). doi: 10.1109/ICMLCN64995.2025.11140453

Efficient 5G Resource Block Scheduling Using Action Branching and Transformer Networks

Abstract

Publication series

Conference

Keywords

Access to Document

Other files and links

Fingerprint

Cite this