Ultra-dense LEO satellite-aircraft access and service management in civil aviation

With the deployment of ultra-dense low earth orbit (LEO) satellite constellations, LEO satellite access network (LEO-SAN) is envisioned to achieve global Internet coverage. Meanwhile, the civil aviation communications have increased dramatically, especially for providing airborne Internet services. However, due to dynamic service demands and onboard LEO resources over time and space, it poses huge challenges in satellite-aircraft access and service management in ultra-dense LEO satellite networks (UDLSN). In this paper, we propose a deep reinforcement learning-based approach for ultra-dense LEO satellite-aircraft access and service management. Firstly, we develop an airborne Internet architecture based on UDLSN and design a management mechanism including medium earth orbit satellites to guarantee lightweight management. Secondly, considering latency-sensitive and latency-tolerant services, we formulate the problem of satellite-aircraft access and service management for civil aviation to ensure service continuity. Finally, we propose a proximal policy optimization-based access and service management algorithm to solve the formulated problem. Simulation results demonstrate the convergence and effectiveness of the proposed algorithm with satisfying the service continuity when applying to the UDLSN.