A Novel Antenna Tracking Method for LEO Satellites Using Bstar Coefficient Dynamic Calibration

Low Earth Orbit (LEO) satellite communication offers distinct advantages, with precise orbit prediction serving as the foundational requirement for achieving autonomous antenna tracking in these systems. A comparative analysis of the Systems Tool Kit (STK) software and the Simplified General Perturbations 4 (SGP4) model for LEO orbit prediction demonstrates that both methods maintain angular tracking errors below 0.5° in short-term predictions, meeting the accuracy requirements for small wide-beam antenna tracking. Furthermore, the SGP4 model proves more advantageous in terms of cost efficiency and operational flexibility. For medium- to long-term orbit prediction errors, a dynamic optimization framework is proposed. By integrating historical Two Line Elements (TLE) data and space weather parameters, the framework employs a genetic algorithm to optimize the ballistic coefficient globally, minimizing azimuth and elevation prediction errors. Experimental results show that the optimized model improves prediction accuracy by over 12% in both azimuth and elevation angles across a 72-hour forecasting window. Furthermore, experimental validation using a phased-array antenna with a gain of 32 dBi and a half-power beamwidth of 3 ° confirmed the method’s effectiveness. With the beam steered via Ethernet, continuous and stable signal reception was achieved when pointing to the predicted angles. This approach provides a robust and cost-efficient solution for autonomous antenna tracking in LEO satellite communication systems. Our code will be uploaded to https://github.com/miemie-323/SBBP-orbit-code.git after the acceptance of this manuscript.