ENHANCING CUBESAT TELEMETRY SYSTEMS FOR AUTONOMOUS SPACE MISSIONS UTILIZING MACHINE LEARNING TECHNIQUES

Abstract

A CubeSat is a valuable tool used by many organizations, including NASA, who partners with universities to design and build satellites for data collection. A primary challenge for CubeSats is maintaining reliable telemetry during autonomous operations. The objective of this paper is to present a machine learning-driven approach to improve real-time data analysis and anomaly detection. The proposed algorithm has the potential to improve the decision-making and reliability of the CubeSat telemetry system, while addressing its unique constraints. The machine learning algorithm, incorporating data supplied by Attitude Determination and Control System (ADCS) components, could find new avenues to increase the efficiency of satellite reorientation based on supplied attitude determination data. Enhancements to the CubeSat operating system could allow for more effective research of autonomous space missions.