This project introduces a novel demand-driven, heterogeneous, multi-layer UAV system architecture that dynamically
adapts sensing, communication, and computing resources to diverse applications. By integrating unmanned aerial
vehicles (UAVs) with complementary capabilities, the system aims to deliver optimized sensing, data collection, mobile
user connectivity, and edge computing services. The core innovation lies in AI-driven UAV path planning, coordinated
positioning of heterogeneous UAV layers (sensing, communication, computing), and real-time resource allocation
tailored to application-specific requirements. The architecture will be evaluated across multiple scenarios, including
disaster management, environmental monitoring, mobile connectivity, and surveillance. In addition, a comprehensive
simulation environment and a proof-of-concept demonstration will be developed. By bridging theoretical models with
practical implementations, the project contributes to scalable, real-world–ready next-generation non-terrestrial network
(NTN) solutions.
