Wide-band Circular Slot Antenna Beam Switching Directions Using Machine Learning Algorithms

This research presents a wideband beam-switching antenna operating at a frequency of 2.6 GHz, which is one of the frequencies for 5G. The technique employed involves using machine learning to predict the direction of the beam switch based on varying three parameters: the distance from the center point of the circular antenna, the position, and the number of shorted circuit points. The study uses a dataset of 100 different configurations obtained from designs in CST Studio. The predicted directions of the beam switch are then compared using five different algorithms: Random Forest, K-Nearest Neighbors (KNN), Decision Trees, Linear Regression, and Lasso Regression. Among all the algorithms, Decision Trees and Random Forest provided the best results.