Interference Management in ISAC–SAGINs Based on Transformer-Enabled Mean-Field Reinforcement Learning Method
IEEE Transactions on Wireless Communications · 2026 · 25 · 19628-19642
DOI: 10.1109/twc.2026.3707527 ↗Kurz gefasst
Fachartikel in der Zeitschrift IEEE Transactions on Wireless Communications, 2026. Verfasst von Yu Yao und 5 weiteren Autor:innen. Die Arbeit wurde bisher 3-mal zitiert (Crossref) und nennt 37 Literaturangaben. Gefördert von: National Natural Science Foundation of China, Hainan Provincial Natural Science Foundation of China, Fundamental and Interdisciplinary Disciplines Break through Plan of the Ministry of Education of China.
Abstract (Originalsprache)
This paper proposes a novel interference management scheme for integrated sensing and communication (ISAC)-enabled space-air-ground integrated networks (SAGINs), where a collaborative multi-UAV network that coexists with a satellite system. Due to communication overhead and privacy concerns, neither satellite nor UAV users necessarily share a willingness to participate in interference management. We model the satellite-UAV interaction as a non-cooperative Stackelberg game. Specifically, the satellite leader first optimizes its beamforming to maximize achievable rate under power budget constraint; the multi-UAV followers then dynamically allocate resources to maximize their average rate while meeting sensing and trajectory requirements. To tackle the complex hierarchical game problem characterized by initial sensitivity and high computational complexity, we propose a mean-field reinforcement learning (MFRL) algorithm integrated with a transformer encoder architecture. By simplifying multi-agent competition to interactions with the mean field, the transformer-enabled MFRL method becomes suitable for analyzing large-scale scenarios and allows interference to be efficiently calculated via mean field interference, greatly reducing computational complexity. Simulation results show that the proposed algorithm improves the average achievable rate by 54.2%, while maintaining acceptable sensing performance, compared to existing learning-based benchmarks.
Förderung
- National Natural Science Foundation of China
- Hainan Provincial Natural Science Foundation of China
- Fundamental and Interdisciplinary Disciplines Break through Plan of the Ministry of Education of China
- Fundamental Research Funds for the Central Universities, Sun Yat-sen University
Zitieren
APA
Yao, Y., Lu, Z., Chen, G., Huang, C., Feng, C., & Quek, T. Q. S. (2026). Interference Management in ISAC–SAGINs Based on Transformer-Enabled Mean-Field Reinforcement Learning Method. IEEE Transactions on Wireless Communications, 25, 19628-19642. https://doi.org/10.1109/twc.2026.3707527
Vancouver
Yao Y, Lu Z, Chen G, Huang C, Feng C, Quek TQS. Interference Management in ISAC–SAGINs Based on Transformer-Enabled Mean-Field Reinforcement Learning Method. IEEE Transactions on Wireless Communications. 2026;25:19628-19642. doi:10.1109/twc.2026.3707527
BibTeX
@article{Yao2026,
author = {Yao, Yu and Lu, Zekun and Chen, Gaojie and Huang, Chong and Feng, Chenyuan and Quek, Tony Q. S.},
title = {Interference Management in ISAC–SAGINs Based on Transformer-Enabled Mean-Field Reinforcement Learning Method},
journal = {IEEE Transactions on Wireless Communications},
year = {2026},
volume = {25},
pages = {19628--19642},
doi = {10.1109/twc.2026.3707527}
}
EUVIMED