Abstract: Aiming to obtain a low sidelobe of an array, fuzzy discrete particle swarm optimization (FDPSO) algorithm is firstly applied in the optimization of the element positions of thinned antenna arrays. Compared with some evolutionary algorithms such as genetic algorithm, the FDPSO has the advantages of less parameters and easier to be implemented. In order to avoid the algorithm being trapped into local optimum, a chaotic dynamic procedure is introduced in the optimization. A penalty approach is used in the constraint of the width of mainlobe. Simulation results of a linear array show that the algorithm can handle discrete problem properly.