Reconfigurable array design using parallel particle swarm optimization

Abstract

Reconfigurable antenna arrays that are capable of radiating with multiple power patterns using a single power divider network are desirable for many applications. We use the particle swarm optimization (PSO) algorithm to perform dual-beam array optimization. Novel features of the PSO algorithm are its generality and ease of implementation over a wide range of electromagnetic problems. The problem investigated is as follows: design a reconfigurable dual-beam 20 element array such that the same amplitude distribution generates either a pencil-shaped or a sector power pattern, the difference being dependent only upon the phase distribution of the array. All excitation phases are set at 0/spl deg/ for the pencil-shaped beam, and are varied in the range -180/spl deg/