Extremum Seeking Control Techniques for Antenna Pointing

This paper considers the application of extremum-seeking control (ESC) to the problem of ground station antenna pointing to maximize the received power. We compare three ESC algorithms from the literature: batch least square (BLS), recursive least squares with exponential forgetting factor (eRLS), and classical ESC (CESC). The three ESC algorithms are also compared with an existing algorithm from the field conical scan (CONSCAN). We present the models for the ground station dynamics, transmitting satellite motion, and received power developed from the literature. The four algorithms are compared under three different simulation scenarios: a static optimum point, a static optimum point with noise on the received power, and tracking a moving optimum point. The static scenario evaluates algorithm performance under ideal conditions. The perturbation scenario adds noise to the power measurements. Finally, the tracking scenario considers a more realistic scenario where the optimum pointing is changing due to satellite motion.