Study, analysis, design and development of space deployable antenna

Abstract

The present study works to develop advanced antenna designs which resolve growing demands for efficient satellite communications technology. The project solves deployment obstacles as well as weight and storage problems by implementing microstrip patch antennas instead of bulky mechanical constructions. Our research is performed on Ansys HFSS simulations to analyse several patch antennas using several different substrates. We optimized these designs to improve gain levels and return loss results along with multi-frequency capability. A key contribution of this work is the introduction and simulation of novel patch antenna designs, including fractal, symbolic, and stacked configurations, aimed at improving gain and return loss. We also applied several AI & ML models like Random Forest and Decision Tree and Gradient Boosting models formed part of the performance enhancement process that predicted antenna features based on geometrical data. Using ML & AI infrastructure the framework we also designed models capable of computing proper dimensions according to specified frequency requirements. The Vector Network Analyzer tests verified the performance predictions which showed both simulated and AI-generated data produced results that demonstrated effective impedance matching and multi-band operating features for the built antennas. This integrated method enhances the creation of smart deployable antenna systems which can be applied to future CubeSats and satellite models. Modern satellite communication technology benefits from innovative design methods which unite with machine learning optimization to bring essential benefits like automated operations while supporting scale and easy mechanical design for contemporary space applications.