A Dual-Band Millimeter-Wave TT&C Choke-Ring Antenna Developed Using Generative Artificial Intelligence
Authors:
Nelson J. G. Fonseca, Charalampos Stoumpos, Maddy Clements, Alberto Montemiglio, Jonathan Davis, Tomas Navarro Reverte, Nicolas Capet
Abstract:
This work explores the use of artificial intelligence (AI) for the design and optimization of millimeter-wave SatCom antennas. The case study reported here is a dual-band choke-ring antenna, a type of antenna that is being considered for future Low Earth Orbit (LEO) satellite constellations, specifically for the Telemetry, Tracking and Command (TT&C) onboard sub-system. The goal of the study was to evaluate the capability of generative AI (GenAI) techniques to improve upon previous results obtained using conventional optimization techniques. The targeted TT&C application requires a wide beam in two separate bands, with a ratio of approximately 1:1.5. Choke-ring antennas are by design narrow-band and have a high number of design variables, constituting a challenging case for any optimization technique. Despite the limited training data set generated within the study, some improvement in performance has been achieved using a dedicated GenAI tool based on a Variational Autoencoder (VAE). The best AI-generated antenna profile was selected for verification through the manufacture and test of a fully-functional prototype. A good agreement between numerical and experimental results was achieved. The results reported here showcase the potential of GenAI techniques for the optimization of novel antennas with complex and unconventional profiles. Further studies are required to confirm the conclusions of the work, extending the training dataset with more complex geometries, now within reach thanks to emerging additive manufacturing techniques, and with a more integrated and iterative design process.
