Using Global Optimization Approaches to Reconstruct Radiation Patterns

Abstract

The paper deals with the reconstruction method of the radiation pattern of directional antennas. The method is based on the amplitude near-field measurement (or simulation) on two plane surfaces, and consists of two steps. First, a global optimization is used to find an initial estimation. Second, a common Fourier iterative algorithm is created to iteratively improve the initial estimation. The genetic algorithm and the particle swarm optimization were chosen to determine the initial estimation. Their comparison was carried out on the radiation pattern reconstruction of the GSM-900 antenna.