Development of a High Gain Fss Reflector Backed Monopole Antenna Using Machine Learning for 5g Applications

Abstract

—This work is devoted to the development of a high gain Frequency Selective Surface (FSS) reflector backed monopole antenna using Machine Learning (ML) techniques for 5G applications. It analyzes and solves the complexity of the determination of the optimum position of the FSS reflector and the ground dimension of the monopole in this composite antenna structure since there are no solid and standard formulations for the computation of these two parameters. ML modelling is involved in the development process for the sake of gain enhancement. It is applied to get the optimum position of the FSS reflector layer and the ground dimension of the monopole antenna. The proposed antenna structure is 50 mm × 50 mm, implemented on a Rogers 5880 substrate (thickness = 1.6 mm). Two different patch antenna structures, with and without FSS, are developed and considered in the current work. The antenna performance in terms of operating frequency, return loss, and gain is analysed using the finite element methods. The design is optimized for a targeting frequency band operating at 6 GHz (5.53 GHz to 6.36 GHz), which is suitable for 5G Sub-6 GHz applications. The obtained results show that the integration of the FSS layer below the antenna structure provides a simple and efficient method to obtain a low-profile and high-gain antenna. Finally, the proposed design is fabricated and measured, and a good agreement between the simulated and measured results is obtained. A comparison with similar studies in the literature is presented and shows that the current design is more compact in size, and the obtained radiation efficiency and gain are higher than other designs. © 2021, Electromagnetics Academy. All rights reserved.