Resilience-Driven Topology Reconfiguration via Hierarchical Deep Reinforcement Learning in Low-Altitude UAV Networks

Jingbin Zhang; Dezhi Zheng; Zhengzhi Yang; Yumeng Li; Wenbo Du; Tony Q.S. Quek; Shuai Wang

doi:10.1109/WCNCW67598.2026.11555081

Resilience-Driven Topology Reconfiguration via Hierarchical Deep Reinforcement Learning in Low-Altitude UAV Networks

Jingbin Zhang
, Dezhi Zheng
, Zhengzhi Yang
, Yumeng Li
, Wenbo Du
, Tony Q.S. Quek
, Shuai Wang

Advanced Research Institute of Multidisciplinary Science

Beihang University
Singapore University of Technology and Design

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

Abstract

In low-altitude wireless networks, Unmanned Aerial Vehicles (UAVs) are susceptible to environmental disturbances that may disrupt topology and degrade coverage and backhaul performance. To address this, we propose UR-HDRL (UAV network Reconfiguration based on Hierarchical Deep Reinforcement Learning), a hierarchical framework that decouples safety constraints from communication optimization. By integrating Control Barrier Functions (CBFs) and Graph Neural Networks (GNNs), UR-HDRL ensures safety and enhances cooperative decision-making in obstacle-rich environments. Experimental results demonstrate notable improvements in transmission efficiency, coverage, and collision avoidance over baseline methods.

Original language	English
Title of host publication	2026 IEEE Wireless Communications and Networking Conference Workshops, WCNCW 2026
Publisher	Institute of Electrical and Electronics Engineers Inc.
ISBN (Electronic)	9798331577315
DOIs	https://doi.org/10.1109/WCNCW67598.2026.11555081
Publication status	Published - 2026
Event	2026 IEEE Wireless Communications and Networking Conference Workshops, WCNCW 2026 - Kuala Lumpur, Malaysia Duration: 13 Apr 2026 → 16 Apr 2026

Publication series

Name	2026 IEEE Wireless Communications and Networking Conference Workshops, WCNCW 2026

Conference

Conference	2026 IEEE Wireless Communications and Networking Conference Workshops, WCNCW 2026
Country/Territory	Malaysia
City	Kuala Lumpur
Period	13/04/26 → 16/04/26

Keywords

Low-altitude Wireless Networks
Multi-agent Reinforcement Learning
Network Reconfiguration
UAV Communications

Access to Document

10.1109/WCNCW67598.2026.11555081

Cite this

Zhang, J., Zheng, D., Yang, Z., Li, Y., Du, W., Quek, T. Q. S., & Wang, S. (2026). Resilience-Driven Topology Reconfiguration via Hierarchical Deep Reinforcement Learning in Low-Altitude UAV Networks. In 2026 IEEE Wireless Communications and Networking Conference Workshops, WCNCW 2026 (2026 IEEE Wireless Communications and Networking Conference Workshops, WCNCW 2026). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/WCNCW67598.2026.11555081

Zhang, Jingbin ; Zheng, Dezhi ; Yang, Zhengzhi et al. / Resilience-Driven Topology Reconfiguration via Hierarchical Deep Reinforcement Learning in Low-Altitude UAV Networks. 2026 IEEE Wireless Communications and Networking Conference Workshops, WCNCW 2026. Institute of Electrical and Electronics Engineers Inc., 2026. (2026 IEEE Wireless Communications and Networking Conference Workshops, WCNCW 2026).

@inproceedings{0c5bc04a9b0241f3b4dd7afd41423ba9,

title = "Resilience-Driven Topology Reconfiguration via Hierarchical Deep Reinforcement Learning in Low-Altitude UAV Networks",

abstract = "In low-altitude wireless networks, Unmanned Aerial Vehicles (UAVs) are susceptible to environmental disturbances that may disrupt topology and degrade coverage and backhaul performance. To address this, we propose UR-HDRL (UAV network Reconfiguration based on Hierarchical Deep Reinforcement Learning), a hierarchical framework that decouples safety constraints from communication optimization. By integrating Control Barrier Functions (CBFs) and Graph Neural Networks (GNNs), UR-HDRL ensures safety and enhances cooperative decision-making in obstacle-rich environments. Experimental results demonstrate notable improvements in transmission efficiency, coverage, and collision avoidance over baseline methods.",

keywords = "Low-altitude Wireless Networks, Multi-agent Reinforcement Learning, Network Reconfiguration, UAV Communications",

author = "Jingbin Zhang and Dezhi Zheng and Zhengzhi Yang and Yumeng Li and Wenbo Du and Quek, \{Tony Q.S.\} and Shuai Wang",

note = "Publisher Copyright: {\textcopyright} 2026 IEEE.; 2026 IEEE Wireless Communications and Networking Conference Workshops, WCNCW 2026 ; Conference date: 13-04-2026 Through 16-04-2026",

year = "2026",

doi = "10.1109/WCNCW67598.2026.11555081",

language = "English",

series = "2026 IEEE Wireless Communications and Networking Conference Workshops, WCNCW 2026",

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

booktitle = "2026 IEEE Wireless Communications and Networking Conference Workshops, WCNCW 2026",

address = "United States",

}

Zhang, J, Zheng, D, Yang, Z, Li, Y, Du, W, Quek, TQS & Wang, S 2026, Resilience-Driven Topology Reconfiguration via Hierarchical Deep Reinforcement Learning in Low-Altitude UAV Networks. in 2026 IEEE Wireless Communications and Networking Conference Workshops, WCNCW 2026. 2026 IEEE Wireless Communications and Networking Conference Workshops, WCNCW 2026, Institute of Electrical and Electronics Engineers Inc., 2026 IEEE Wireless Communications and Networking Conference Workshops, WCNCW 2026, Kuala Lumpur, Malaysia, 13/04/26. https://doi.org/10.1109/WCNCW67598.2026.11555081

Resilience-Driven Topology Reconfiguration via Hierarchical Deep Reinforcement Learning in Low-Altitude UAV Networks. / Zhang, Jingbin; Zheng, Dezhi; Yang, Zhengzhi et al.
2026 IEEE Wireless Communications and Networking Conference Workshops, WCNCW 2026. Institute of Electrical and Electronics Engineers Inc., 2026. (2026 IEEE Wireless Communications and Networking Conference Workshops, WCNCW 2026).

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

TY - GEN

T1 - Resilience-Driven Topology Reconfiguration via Hierarchical Deep Reinforcement Learning in Low-Altitude UAV Networks

AU - Zhang, Jingbin

AU - Zheng, Dezhi

AU - Yang, Zhengzhi

AU - Li, Yumeng

AU - Du, Wenbo

AU - Quek, Tony Q.S.

AU - Wang, Shuai

PY - 2026

Y1 - 2026

N2 - In low-altitude wireless networks, Unmanned Aerial Vehicles (UAVs) are susceptible to environmental disturbances that may disrupt topology and degrade coverage and backhaul performance. To address this, we propose UR-HDRL (UAV network Reconfiguration based on Hierarchical Deep Reinforcement Learning), a hierarchical framework that decouples safety constraints from communication optimization. By integrating Control Barrier Functions (CBFs) and Graph Neural Networks (GNNs), UR-HDRL ensures safety and enhances cooperative decision-making in obstacle-rich environments. Experimental results demonstrate notable improvements in transmission efficiency, coverage, and collision avoidance over baseline methods.

AB - In low-altitude wireless networks, Unmanned Aerial Vehicles (UAVs) are susceptible to environmental disturbances that may disrupt topology and degrade coverage and backhaul performance. To address this, we propose UR-HDRL (UAV network Reconfiguration based on Hierarchical Deep Reinforcement Learning), a hierarchical framework that decouples safety constraints from communication optimization. By integrating Control Barrier Functions (CBFs) and Graph Neural Networks (GNNs), UR-HDRL ensures safety and enhances cooperative decision-making in obstacle-rich environments. Experimental results demonstrate notable improvements in transmission efficiency, coverage, and collision avoidance over baseline methods.

KW - Low-altitude Wireless Networks

KW - Multi-agent Reinforcement Learning

KW - Network Reconfiguration

KW - UAV Communications

UR - https://www.scopus.com/pages/publications/105043318574

U2 - 10.1109/WCNCW67598.2026.11555081

DO - 10.1109/WCNCW67598.2026.11555081

M3 - Conference contribution

AN - SCOPUS:105043318574

T3 - 2026 IEEE Wireless Communications and Networking Conference Workshops, WCNCW 2026

BT - 2026 IEEE Wireless Communications and Networking Conference Workshops, WCNCW 2026

PB - Institute of Electrical and Electronics Engineers Inc.

T2 - 2026 IEEE Wireless Communications and Networking Conference Workshops, WCNCW 2026

Y2 - 13 April 2026 through 16 April 2026

ER -

Zhang J, Zheng D, Yang Z, Li Y, Du W, Quek TQS et al. Resilience-Driven Topology Reconfiguration via Hierarchical Deep Reinforcement Learning in Low-Altitude UAV Networks. In 2026 IEEE Wireless Communications and Networking Conference Workshops, WCNCW 2026. Institute of Electrical and Electronics Engineers Inc. 2026. (2026 IEEE Wireless Communications and Networking Conference Workshops, WCNCW 2026). doi: 10.1109/WCNCW67598.2026.11555081

Resilience-Driven Topology Reconfiguration via Hierarchical Deep Reinforcement Learning in Low-Altitude UAV Networks

Abstract

Publication series

Conference

Keywords

Access to Document

Other files and links

Fingerprint

Cite this