Design, Implementation and Machine Learning Analysis of Frequency Reconfigurable Microstrip Antenna for Defense Applications

Abstract

In this paper a Frequency Reconfigurable antenna with a U-slot is designed with a high gain of 9.2dB and can be reconfigured from 1.61-1.68GHz, the analysis of reconfigurability aspect and the behaviour of lumped components is studied using Machine Learning. Reconfigurable antennas are those which are capable of changing the resonant frequency based on the switching circuits used. These switching circuits use an additional load such as PIN Diodes, MEMS Switches etc., We considered PIN Diode since it is easy to design and is economical for our analysis in HFS S. The effect of each lumped component (R, L &C) is individually studied and it is observed that R value has a correlation coefficient of 0.961 with return loss, and correlation coefficient of 0.090 is obtained for frequency. R value is said to have significant effect on reconfigurability aspect of the antenna.