APPLYING MACHINE LEARNING TO THE PREDICTION OF OUTPUT PARAMETERS IN ANTENNA DESIGN

Abstract

This paper presents the application of machine learning (ML) to predict and optimize parameters in antenna design. The main contribution of the paper is using a ML model with training and evaluation dataset as parameters obtained from antenna simulation results by CST software. The results of the model are used to predict antenna dimensions at the desired resonant frequency. In this paper, the K-Nearest Neighbors (KNN) algorithm is applied to predict the parameters of the antenna patch. The accuracy of the prediction results is evaluated and analyzed using root mean square error (RMSE). These results can provide a basis and new direction to improve the antenna design process, contributing to progress in deploying more modern and efficient wireless systems. The prediction results contribute to reducing the time to optimize parameters in the antenna design.