Analysis of Resonant Characteristics and Radiation Patterns of a Circular Microstrip Antenna on Isotropic or Uniaxially Anisotropic Substrate Using Artificial Neural Network

Abstract

In this work a single neural network for determining the resonant frequency, quality factor and far-zone radiation patterns of a circular disk microstrip antenna, printed on a uniaxial anisotropic substrate using effective parameters in conjunction with spectral dyadic Green’s function. The resonant frequency results predicted by the numerical results show that there are substantial deviations in calculated resonant frequency and quality factor when substrate dielectric anisotropy is considered. Furthermore, significant variations are seen in the radiation patterns of the structures due to substrate anisotropy. The variations of resonant frequency, quality factor and radiation patterns of the structure, with respect to anisotropy ratio of the substrate, for several values of substrate thickness and patch radius are presented.