Satellite Antenna Design Using AI Based Approach

This study explores an AI-driven neurocomputational analysis method for the analysis and performance optimization of microstrip patch antennas with versatile satellite frequency bands. A specialized artificial neural network was developed and trained to process geometrical design variables such as patch length, width, and feed point location as inputs. This trained network determined the resonance frequency and bandwidth of microstrip patch antennas based on the provided parameters. The credibility of this neural network model is evaluated by comparing its performance against simulation data generated from CST Software, along with experimental findings obtained from a few fabricated prototype antennas.