Artificial Neural Network Model for Calculating the Dimensions of Circular and Semi-Circular Microstrip Antennas for Given Resonance Frequency

Since the invention of microstrip antenna in 1969, many advances have taken place in their design and implementation. Ideally in microstrip antenna, any shape for the radiating patch can be used, but from the numerical modeling point of view, regular shape geometries have been commonly employed. To evaluate the resonance frequency for the given radiating patch shape, frequency equation for the regular geometries has been reported. However, they fail to predict the frequency accurately on the varying thickness of the substrate and also over wide frequency range. This paper presents the artificial neural network model for predicting the radius of the patch for the given frequency for circular and semi-circular microstrip antenna on air as well as the suspended dielectric thicker substrates. Using proposed model, the prediction in patch radius has been done across wide frequency range and for the thickness of the substrate in the range of 0.02–0.1λg. For all these ranges, close prediction in desired frequency values has been realized.