Aerial Base Station enabled the Rocket Communication System

Non-Terrestrial Networks (NTNs) play a critical role in ensuring uninterrupted communication services. Reliable communication between launch vehicles and ground stations is a cornerstone of mission-critical telemetry, tracking, and command (TT&C) systems. During high-dynamic phases such as ascent, main engine cutoff, serial staging, and re-entry burns, traditional ground-based antennas often encounter reduced signal quality due to rapidly changing elevation angles, Doppler effects, and terrain obstructions. These effects are exacerbated over oceanic re-entry zones, where terrestrial coverage is limited. To address this issue, this study proposes the integration of aerial vehicles (AVs) as the aerial segment, forming an Air-Ground Integrated Network (AGIN). AVs leverage their flexibility and ability to transmit via line-of-sight (LoS), act as wireless relay stations, effectively enhancing rocket signal transmission and improving Quality of Service (QoS), and agile relaying of communication signals. In this work, we explore a low-based airborne relay system capable of dynamically adjusting its position to maintain a reliable communication link with the rocket and forward the signal to a terrestrial base station (TBS). Next, we employ Deep Reinforcement Learning (DRL) techniques to optimize the UAV 3D position. Simulation results demonstrate that this approach significantly improves the network communication efficiency, providing essential theoretical foundations and technical references for achieving reliable and efficient SAGIN systems.