Artificial neural network based design of a three-layered microstrip circular ring antenna with specified multi-frequency operation

Abstract

This paper deals with the utilization of Artificial Neural Networks (ANN) methodology in the design of microstrip antennas having a priori defined operational features. A printed annular ring antenna, textured by slits, was designed in a multilayered dielectric substrate. The width and the position of the slits along with the values of the structural parameters and the feed position of the antenna affect directly its frequency performance. The inverse problem, namely to find out the proper combination of all these parameter values which would yield the desired frequency response, was solved via a Multiple Layer Perceptron (MLP) Neural Network.