EUVIMEDEUVIMED entdecken
Crossref

Covert Semantic Communication in UAV Networks: A Cross-Layer Dual-Attention DRL Framework for Finite Blocklength Optimization

Xinfeng Deng, Li Zhou, Zhe Wang, Nan Li, Haitao Zhao, Haijun Zhang, Dongtang Ma

IEEE Transactions on Wireless Communications · 2026 · 25 · 25490-25503

DOI: 10.1109/twc.2026.3743250 ↗

Kurz gefasst

Fachartikel in der Zeitschrift IEEE Transactions on Wireless Communications, 2026. Verfasst von Xinfeng Deng und 6 weiteren Autor:innen. Die Arbeit nennt 41 Literaturangaben. Gefördert von: National Natural Science Foundation of China.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Open wireless channels in uncrewed aerial vehicle (UAV) networks pose significant security challenges because traditional security techniques do not conceal the presence of communication. While covert communication mitigates this issue, it faces throughput limitations in high-capacity transmission scenarios. This paper proposes a finite blocklength covert semantic communication framework for multi-UAV networks that integrates communication, computation, control, and semantics. We first define two cross-layer metrics, cross-layer covert semantic throughput (CCST) and cross-layer covert semantic rate (CCSR), which map task-level semantic requirements to physical-layer performance. We further derive a theoretical upper bound on CCST, revealing a fundamental trade-off between semantic accuracy and covertness governed by the square-root law. To optimize covert semantic communication under finite blocklength constraints, we design a dual-attention-enhanced QMIX (DA-QMIX) scheme that learns robust policies by adaptively attending to environmental features. Furthermore, we develop a multi-dimensional joint optimization framework that dynamically optimizes user scheduling, trajectory planning, and blocklength selection. Simulation results show that the proposed DA-QMIX scheme improves CCST by 8% over graph neural network-based schemes and by 37% over standard QMIX. The framework enables UAVs to achieve low-latency covert transmission while preserving semantic task performance under finite blocklength constraints.

Förderung

  • National Natural Science Foundation of China

Zitieren

APA

Deng, X., Zhou, L., Wang, Z., Li, N., Zhao, H., Zhang, H., & Ma, D. (2026). Covert Semantic Communication in UAV Networks: A Cross-Layer Dual-Attention DRL Framework for Finite Blocklength Optimization. IEEE Transactions on Wireless Communications, 25, 25490-25503. https://doi.org/10.1109/twc.2026.3743250

Vancouver

Deng X, Zhou L, Wang Z, Li N, Zhao H, Zhang H, et al. Covert Semantic Communication in UAV Networks: A Cross-Layer Dual-Attention DRL Framework for Finite Blocklength Optimization. IEEE Transactions on Wireless Communications. 2026;25:25490-25503. doi:10.1109/twc.2026.3743250

BibTeX
@article{Deng2026,
  author = {Deng, Xinfeng and Zhou, Li and Wang, Zhe and Li, Nan and Zhao, Haitao and Zhang, Haijun and Ma, Dongtang},
  title = {Covert Semantic Communication in UAV Networks: A Cross-Layer Dual-Attention DRL Framework for Finite Blocklength Optimization},
  journal = {IEEE Transactions on Wireless Communications},
  year = {2026},
  volume = {25},
  pages = {25490--25503},
  doi = {10.1109/twc.2026.3743250}
}
Kostenlos starten →