Analysis of Circular Antenna Array Using Machine Learning Technique

This paper presents an approach to design 20, 36-elements asymmetric circular antenna array (CAA) for optimum far-field radiation pattern synthesis. By minimizing the sidelobe levels (SLLs), the far-field radiation pattern is designed to minimize the sidelobe level and first null beamwidth of CAA. For achieving the ultra-low SLLs, an optimal combination of excitation amplitude and phase is determined using a proposed machine learning technique which is Gradient Decent.