Dual port ring cylindrical dielectric resonator antenna optimization using ML algorithm

Abstract

In the present scenario, Machine Learning techniques are used in much ongoing research as a powerful tool. This paper proposes the applications of machine learning in antenna design optimization by implementing different machine learning algorithms like KNN, ANN, Random Forest, XGBoost and Decision Tree. A Double ring Cylindrical Di-electric Resonator Antenna is designed using High-Frequency Structure Simulator (HFSS). For the proposed antenna design, the frequency range is 2–3.5 GHz, while the range of height and radius is 6.5–19.5 mm and 12–18 mm respectively. The data set is generated for the proposed antenna design and S11 parameter is optimized using machine learning algorithms. Out of the five algorithms, the models for KNN, XGBoost and Artificial Neural Network perform almost similarly and Random Forest has the highest performance.