Machine Learning Approaches for Enhancing the Design of a 5G MIMO Antenna

This study explores the application of modern machine learning (ML) techniques to optimize the design of CPW-fed MIMO antennas for operation within specific frequency ranges. Using automated methodologies, the proposed approach provides a robust, adaptable, and efficient framework to determine optimal design parameters. A comprehensive analysis and comparative evaluation of the predictive performance of various ML models is conducted to validate their accuracy. Among the models tested, the best R<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> scores achieved were 0.94153 for Gradient Boosting, 0.98608 for Random Forest, and 0.986474 for K-Nearest Neighbors (KNN). These ML predictions were further validated by comparisons with results obtained from the CST high-frequency structure simulator, demonstrating the effectiveness of the proposed approach.