Metasurface frequency reconfigurable antenna optimizes using neural network algorithm for wireless applications

Abstract

Abstract As the demand increases in the field of wireless communication system, the interest of researchers increases to develop and analyzes the antenna for these applications. This article present metasurface (MS) frequency reconfigurable antenna (FRA) which is optimized using neural network (NN) approach. The designed structure is a multilayer consists of three-layer patch antenna placed under the MS structure. The resonating patch is a rectangular shaped and substrate is a circular shaped show as to reduce the geometry of the antenna. The proposed structure is fabricated on FR-4 substrate of thickness 1.6 mm. The MS structure consists of split ring rectangular (SRR) strips made up of copper. The antenna reconfigured the operating frequency from 4.85 to 7 GHz having overall bandwidth of 2.15 GHz with wide range of tuning. The central frequency of rectangular patch antenna is 6.15 GHz. The MS is analyzed by using effective parameters i.e., effective permittivity (ε r) & effective permeability (μ r) and it is observed that the MS is behaving as a metamaterial in the desired range of frequency. The reconfigured operating frequency (ROF) is found at the anticlockwise rotation angles of 0°, 30°, 600 and 90°. The realized gain and radiation efficiency are calculated at each ROF. The validation of proposed MS based FRA is carried out first by simulating using Ansys HFSS and then measured inside the anechoic chamber. The proposed antenna is optimizes using NN model which shows minimum error during analysis and synthesis process.