A2.3 - Link Resource SchedulingTechnology of Networked Telemetry System for LEO Constellations
- Event
- ETTC 2024 - European Test and Telemetry Conference
2024-06-11 - 2024-06-13
Nuremberg - Chapter
- ML & AI Session I
- Author(s)
- J. Hao, J. Zhang, Y. Wang - Beijing Institute of Astronautical Systems Engineering, Beijing (China)
- Pages
- 40 - 46
- DOI
- 10.5162/ETTC2024/A2.3
- ISBN
- 978-3-910600-02-7
- Price
- free
Abstract
In recent years, many mega Low Earth Orbit (LEO) satellite constellation plans to provide global Internet services have emerged internationally. The Tracking, Telemetry and Command (TT&C) system is the life line to ensure the normal operation of LEO satellites, and the telemetry systemis the key in monitoring the operation status of LEO satellites. The networked telemetry system which is based on High Earth Orbit (HEO) and Medium Earth Orbit (MEO) satellites has no coverage blind area, is applicable to various types of LEO constellations and can serve LEO satellites with abnormal attitudes. Concerning the stable-burst hybrid transmission requirements of LEO satellite telemetry data, we studies the distributed scheduling algorithm of ISL resources in the networked telemetry system. Based on the multi-agent deep reinforcement learning architecture, each MEO satellite dynamically allocates MEO beam and time slot resources according to the latest state of the system, and prioritizes the transmission of high-valuedata to increase the value of the total transmitted data of the system. Meanwhile, we designs a deep Q-network architecture to reduce network traning and storage overhead. The simulation results show that the value of the total telemetry data transmitted by the system is not less than 85% ofthe upper bound.