Aperture coupled dielectric resonator antenna optimisation using machine learning techniques

Abstract

In this article, various machine learning (ML) algorithms such as Artificial neural network (ANN), Random Forest, XG boost, K nearest neighbor (KNN), and Knowledge-based neural network (KBNN) are used for efficient optimization of dielectric resonator antenna (DRA). ML models are used to predict the |S11| for a particular set of frequency, resonator height, aperture radius, and resonator radius. Finally, a comparative performance analysis of different ML algorithms has been done with the outcomes of the HFSS EM simulator. Error percentage is in between 1.0% and 7.0% with different ML algorithms. Antenna design is also fabricated and tested. The performance of the fabricated prototype is very close to different ML algorithms and HFSS obtained outcomes.