A memetic particle swarm optimization algorithm (MPSO) is applied to fine-tune the asymmetrical difference pattern of a linear array, which is useful for tracking targets, for example, in radar applications. The side-lobe level of the asymmetrical difference pattern with various peak differences has been successfully reduced, while maintaining the intended squint angle and the side-lobe difference. Conventional PSO, genetic algorithm (GA) and memetic GA (MGA) have also been applied to uniformly-excited and Bayliss linear arrays to compare their performance of optimization.