Estimation of Scattering Parameters of U-Slotted Rectangular RFID Patch Antenna with Machine Learning Models

Authors : Ismail AKDAG

Pages : 63-70

View : 17 | Download : 12

Publication Date : 2021-08-30

Article Type : Research Paper

Abstract :In this study, machine learning-based models have been used to estimate the return loss parameters of the operational resonant frequency of the U-slotted UHF RFID antenna. The data set utilized, consisting of 544 instances, has been collected from the simulation software as a consequence of the parametric evaluation of the antenna design parameters. Distinct machine learning methods have been used on two different types of output data, complex and linear scattering parameters, and the models` prediction performance has been evaluated. In the single-output regression models, a mean-square error value of 0.25% with an R2 value of 95.54% was obtained with the Random Forest regression model, and a mean-square error value of 0.85% has been obtained with an R2 value of 91.32% in the multiple-output regression technique.
Keywords : Machine learning, radio frequency identification, regression, RFID, scattering parameters