Using electrically large antennas for array applications may cause grating lobes and this can change the array pattern. To achieve a low side lobe level for boresight pattern for linear or planar array antennas, in the literature it is seen that the common method is the use of feeding each antenna with some amplitude distribution and keeping the distance between the antennas less than a wavelength. However, in this study the distance between the antennas is chosen to be greater than a wavelength. Therefore, the common methods cannot be used. To achieve a low side lob level in the radiation pattern for an array of antennas composed of antennas with a distance greater than one wavelength. The aperture length of each antenna is chosen to be variable and an analytical solution method is used to calculate the amplitude of each aperture field of the horn antennas. Thus, the determination of unit horns' optimal aperture lengths Li can be considered a multivariable optimization problem. This challenging problem was achieved using a novel meta-heuristic optimization algorithm called the differential evolutionary algorithm (DEA), where the aperture lengths of each array element are taken as inputs to find the optimal side lobe level (SLL). Based on the obtained solutions, a 4 x 1 horn array antenna design alongside its feed network was designed for Ku-band applications. The design of the array is simulated in CST simulation software, and the obtained simulated results had been compared with experimental results and counterpart designs in the literature.