Miniature three-element stochastic Yagi-Uda array optimization via particle swarm intelligence

The relatively new nature-based algorithm, particle swarm optimization, is used to design optimal miniaturized stochastic Yagi-Uda arrays. The PSO scheme utilizes a fixed-length grid mapping of continuous values. In the conventional sense, the driven linear dipole element of a Yagi-Uda array has a fixed-length at the resonant frequency while the element spacing is known to have the greatest impact on the radiation properties of the antenna. To accomplish miniaturization of such arrays, a simple but effective grid searching technique is introduced for use in conjunction with the PSO. Designs utilizing both parallel and planar grids are considered. It is found that the overall array length and width can be reduced by as much as 70% and 44%, respectively, compared to conventional Yagis. These results are achieved at the expense of some performance parameters, such as directive gain and/or bandwidth.