Levenberg-Marquardt based ANN for design of Rectangular Dielectric Resonator Antenna for LTE Application

Abstract

Antenna design process requires Electromagnetic (EM) simulations which can be performed on EM simulators such as HFSS, CST, ADS and IE3D etc. To solve the antenna design problems, these EM simulators required large computational resources and time. With the increase of parameters and design complexity, simulation cost and time of EM simulators escalates. To overcome this difficulty, Artificial Neural Network (ANN) can be used as an alternative approach for antenna design which greatly reduces computational cost and time. A design of rectangular dielectric resonator antenna (RDRA) based on Artificial neural network approach is presented in this paper for LTE applications. The rectangular resonator having relative permittivity of 30 is placed on top of substrate which has relative permittivity (∊r) of 4.6 and 1.6mm of thickness and simulated by using well-known 3-D electromagnetic (EM) simulator ANSYS HFSS. ANN used consists of one input layer, one hidden layer and one output layer. The neural network is trained using Levenberg-Marquardt algorithm and the data set is divided into 70%, 15% and 15% for training, testing, and validation respectively. The error, described by the difference between the target data and expected output, is 0.007.