In this paper, we propose a general design of rectangular microstrip antenna with and without rectangular aperture over ground plane, based on artificial neural networks (ANN) in conjunction with spectral domain formulation. In the design procedure, syntheses ANN model is used as feed forward network to determine the resonant frequency and bandwidth. Analysis ANN model is used as the reversed of the problem to calculate the antenna and aperture dimensions for the given resonant frequency, dielectric constant and height of substrate. The spectral domain formulation combined with artificial neural network in the analysis and the design of rectangular antenna to reduce the complexity of the spectral approach and to minimize the CPU time necessary to obtain the numerical results. The results obtained from the neural models are in very good agreement with the experimental results available in the literature.

In this paper, the cavity model for simple circular disc microstrip antenna is extended with some modifications for the tunable geometry taking into account the anisotropy in the layer. The numerical results show that there are substantial deviations in calculated resonant frequency when substrate dielectric anisotropy is considered. Furthermore, significant variations are seen in the radiation patterns of the structures due to substrate anisotropy. Finally the effect of inclusion of air gap layer inserted between substrate and ground plane on the resonant characteristics is also investigated for fundamental and higher order modes.