Dynamic Mask Enhanced Intelligent Multi-UAV Deployment for Urban Vehicular Networks

Gaoxiang Cao; Wenke Yuan; Yunpeng Hou; Huasen He; Quan Zheng; Jian Yang

cs.NIMar 19, 2026

Dynamic Mask Enhanced Intelligent Multi-UAV Deployment for Urban Vehicular Networks

Authors: Gaoxiang Cao, Wenke Yuan, Yunpeng Hou, Huasen He, Quan Zheng, Jian Yang

Organizations: Department of Automation, University of Science and Technology of China, Hefei, China · Institute of Artificial Intelligence, Hefei Comprehensive National Science Center, Hefei, China

Abstract

Vehicular Ad Hoc Networks (VANETs) play a crucial role in realizing vehicle-road collaboration and intelligent transportation. However, urban VANETs often face challenges such as frequent link disconnections and subnet fragmentation, which hinder reliable connectivity. To address these issues, we dynamically deploy multiple Unmanned Aerial Vehicles (UAVs) as communication relays to enhance VANET. A novel Score based Dynamic Action Mask enhanced QMIX algorithm (Q-SDAM) is proposed for multi-UAV deployment, which maximizes vehicle connectivity while minimizing multi-UAV energy consumption. Specifically, we design a score-based dynamic action mask mechanism to guide UAV agents in exploring large action spaces, accelerate the learning process and enhance optimization performance. The practicality of Q-SDAM is validated using real-world datasets. We show that Q-SDAM improves connectivity by 18.2% while reducing energy consumption by 66.6% compared with existing algorithms.

Figures & tables

Fig. 1: Multi-UAV Enhanced VANET Scenario

Panel of Figure 2 — Fig. 2: Illustration of RTG and DRTG

f(a_{u}(t))=\begin{cases}-1-l(a_{u}(t),a_{u}(t-1))/\varphi_{0},&\text{if}\ con(a_{u}(t))=1\\ -\dfrac{\alpha_{1}}{\varphi_{0}}l(a_{u}(t),a_{u}(t-1))+\dfrac{\beta_{1}}{\varphi_{1}}\sum\limits_{v\in N(a_{u}(t))}con(v),&\text{otherwise}\end{cases}

(8)

Table 3

Fig. 5: Converge of Intelligent Algorithms

Panel of Figure 7 — Fig. 6: Performance Comparison of Algorithms

Panel of Figure 8 — Fig. 7: Performance Comparison with Varying Number of UAVs

Cooperative MARL

Cooperative multi-agent reinforcement learning (MARL) trains agents to coordinate toward shared goals under partial information and decentralized control.

See all 242 papers

Bridging Network Fragmentation: A Semantic-Augmented DRL Framework for UAV-aided VANETs
Mar 19, 2026Gaoxiang Cao, Wenke Yuan, Huasen He +4Intelligent Transportation Systems Wireless Communications
Department of Automation, University of Science & Technology of China
TRUAV: Distributed Multi-Agent Reinforcement Learning for Trajectory Planning and Routing Enhancement in UAV-Aided IoT-Enabled VANETs
Jul 26, 2026Muhammad Umar Farooq Qaisar, Lin Zhang, Zhen Chen +3UAV Navigation Decentralized MARL
Hangzhou International Innovation Institute of Beihang University, Hangzhou 311115, China · Hangzhou International Innovation Institute of Beihang University, Hangzhou 311115, China, the School of Automation Science and Electrical Engineering at Beihang University, Beijing 100191, China, and the State Key Laboratory of Intelligent Manufacturing Systems Technology, Beijing 100854, China · School of Intelligent Science and Engineering and the School of Transportation, Southeast University, Nanjing 211189, China +3
Hierarchical Multi-Task Federated Learning in VANETs
Aug 8, 2026M. Saeid HaghighiFard, Sinem ColeriHierarchical Federated Learning Communication-Efficient Distributed Training
Department of Electrical and Electronics Engineering, Koc University, Istanbul, Turkey

Mar 19, 2026J/K move · Enter open · S save

Bridging Network Fragmentation: A Semantic-Augmented DRL Framework for UAV-aided VANETs

Gaoxiang Cao, Wenke Yuan, Huasen He, Yunpeng Hou+3

Department of Automation, University of Science & Technology of China

Intelligent Transportation Systems Wireless Communications V2X Communication LLM-Guided RL Proximal Policy Optimization

Open paper AskPDF arXiv

Urban Vehicular Ad-Hoc Networks (VANETs) can become fragmented because buildings obstruct wireless links and vehicle mobility continuously changes the network topology. Unmanned Aerial Vehicles (UAVs) can serve as mobile relays, but Deep Reinforcement Learning (DRL)-based deployment often suffers from inefficient exploration because it lacks road-topology guidance. To address this problem, we propose Semantic-Augmented DRL (SA-DRL), which models network fragmentation over the road topology and aligns a pretrained Large Language Model (LLM) to generate a topology-dependent action prior from dynamic traffic states. The resulting Semantic-Augmented PPO (SA-PPO) algorithm combines this prior with the PPO policy through Logit Fusion, guiding exploration toward promising intersections while retaining adaptation through environmental returns. Simulations driven by real-world urban trajectories show that SA-PPO reaches the final converged reward of Vanilla PPO using only 28.6% of its training episodes. It improves the average number of vehicles in connected components and the average connected-component size by 7.9% and 8.7%, respectively, while reducing UAV energy consumption by 21.3%.

Abstract

Cooperative MARL

Explore similar work

Bridging Network Fragmentation: A Semantic-Augmented DRL Framework for UAV-aided VANETs

TRUAV: Distributed Multi-Agent Reinforcement Learning for Trajectory Planning and Routing Enhancement in UAV-Aided IoT-Enabled VANETs

Hierarchical Multi-Task Federated Learning in VANETs