Deciphering Single Unit Split Ring Resonator and Validation Through Machine Learning Technique

An attempt is made in this article for the design analysis of a single Split Ring Resonator mathematically along with simulations using full wave electromagnetic solver, further it is validated using Supervised Machine Learning Random Forest model. Single unit SRR is proposed at 4.2 GHz frequency with 1 mm gap, 1 mm width and 3 mm height for the transmission coefficient value of −58.2 dB. Various dimensional analysis along width, radius, and ring gap are carried out for resonance frequency. It is SRR gap playing crucial role in resonance frequency due to its gap capacitance. Field distributions and surface currents are obtained. Supervised random forest is applied for its validation by generating the data set from simulator and model validation is carried by Root Mean Square Error Value (RMSE). Feature importance is also obtained on SRR dimensions for various ranges of datasets for training and test models. The reported RMSE for train and test models are 0.014 & 0. 0 2 0 9 respectively, further SRR gap is 76% dominant while determining the resonance frequency.