Scholarlink

This study presents LightGNN, a Graph Neural Network-based framework for optimizing Unmanned Aerial Vehicle (UAV) path planning using personalized recommendations. By integrating user preferences, environmental constraints, and real-time network conditions, LightGNN dynamically adapts UAV routes to enhance efficiency and adaptability. Experimental results show that LightGNN achieves a 14.4% shorter path than Dijkstra while reducing energy consumption by 33.3% and improving signal strength by 17.3%. Additionally, it maintains a computational time of 1.2 s, making it suitable for real-time deployment. Despite scalability and real-world deployment challenges, future work will focus on federated learning and large-scale validation. This research highlights the potential of AI-driven personalized UAV navigation, paving the way for efficient and adaptive aerial systems.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

UAV Path Optimization Using LightGNN-Based Personalized Recommendations

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Luong Vuong Nguyen

摘要

This study presents LightGNN, a Graph Neural Network-based framework for optimizing Unmanned Aerial Vehicle (UAV) path planning using personalized recommendations. By integrating user preferences, environmental constraints, and real-time network conditions, LightGNN dynamically adapts UAV routes to enhance efficiency and adaptability. Experimental results show that LightGNN achieves a 14.4% shorter path than Dijkstra while reducing energy consumption by 33.3% and improving signal strength by 17.3%. Additionally, it maintains a computational time of 1.2 s, making it suitable for real-time deployment. Despite scalability and real-world deployment challenges, future work will focus on federated learning and large-scale validation. This research highlights the potential of AI-driven personalized UAV navigation, paving the way for efficient and adaptive aerial systems.