WiFi Antenna Design and Modeling using Artificial Neural Networks

Abstract

Artificial neural networks (ANN) have gained popularity in microwave modeling, design and simulations. This article is devoted to designing efficient antenna to achieve high gain and optimal impedance matching in addition to employing ANN to model the microstrip antenna. The proposed elliptical patch antenna operates at 2.4 GHz used for wireless applications. The ANN is fed with data set derived by CST EM simulator to train and test the NN model. The feed-forward back-propagation ANN is used along with Levenberg-Marquart (LM) learning algorithm to model the antenna. Extensive analyses has been carried out to provide an efficient ANN model by the aid of statistical measures as mean square error (MSE), mean error and standard deviation error. Moreover, the proposed antenna is fabricated and measured. A high agreement between simulated and measured antenna return loss is illustrated.