Diffusion-Based Trajectory Optimization and Resource Allocation for Hybrid Aircraft in Low-Altitude Economy Networks

This is the authors' abstract. We don't add key points for this paper.

Relying on low-altitude airspace and emerging communication technologies, the low-altitude economy plays a dominant role in the general aviation domain. Robust and efficient low-altitude economy networks with hybrid aircraft are indispensable for supporting complex and large-scale operations of low-altitude economy. However, hybrid aircraft features dynamic spatial structures and diverse task requirements. In this work, we propose a low-altitude economy network framework to support the robust and efficient operations of heterogeneous unmanned aerial vehicles (UAVs) and electric vertical take-off and landing vehicles (eVTOLs). To ensure reliable flights, a secure distance between two aircraft is defined. In the proposed network framework, we formulate a long-term network energy optimization problem for trajectory generation and resource allocation of hybrid aircraft, subject to the constraints of secure distances and task queue stability. Then, the initial time-average energy optimization problem is transformed into deterministic sub-problems across different time slots by Lyapunov optimization. Based on the auxiliary gradients provided by the deterministic sub-problems, a novel diffusion-based trajectory generation (DTG) algorithm is proposed to generate the optimal trajectories of hybrid aircraft with guided sampling, where secure distances act as penalties for trajectory generation. Finally, the effectiveness of the proposed DTG algorithm is evaluated through both theoretical analysis and simulation experiments by comparing it with benchmark policies.