Neural network model for design of compact CPW — fed monopole antenna for 5.8 GHz RFID application

Abstract

In recent years, computer - aided design approach based Artificial Neural Network has been introduced to microwave modeling, simulation & optimization. In this paper, a neural network model is proposed for the design of spiral strip monopole antenna fed by a coplanar waveguide (CPW) for radio frequency identification (RFID) applications is presented. The designed antenna, which, including the FR-4 substrate, with 16.5 mm in length and 20 mm in width, can operate at the frequency of 5.8 GHz and 4.15 dBi gain, respectively. This antenna is suitable for radio frequency identification applications in the 5.8 GHz band. The fundamental parameters of the antenna such as return loss, VSWR, gain and magnitude which meets the acceptable antenna standards are obtained. Simulation tool, based on the method of moments (ZELAND IE3D version 12.0) has been used to analyze and optimize the antenna. This model is prepared using feed forward multilayer perceptron neural network back propagation Levenberg Marqurat algorithm. The neural results are compared with the results of IE3D which shows a very good agreement.