Compact and Efficient modelling of Wideband Semi-Circled U-slot Microstrip Patch Antenna using ML

Abstract

The main objective of this work is to design an efficient wideband Semi-circled U-slot loaded antenna using Machine Learning (ML) algorithm. The proposed concept is for predicting the resonance frequency of the U-slot loaded antenna by providing the dimensions of the antennas. Antenna is designed for WiMAX application with operating frequency band in the range between 0.4856–7.8476 GHz. The HFSS tool is being used for designing and analysing fractal antennas and generating the training data. Parametric analysis of the designed U-slot-loaded half circled antenna is developed by altering the half-circle radius, length of the U-slot and width. The data set is then given to the ANN back-propagation ML algorithm for training the model. The ANN back-propagation contains remarkably high processing speed and contains features like parallelization, cache optimization, and out-of-core computation which makes the perfect algorithm for predicting the resonance frequencies. U-slot loaded half circled antenna offers a substantial size reduction, a wide impedance bandwidth, and a uniform radiation pattern on all sides.