Novel Clustered Subarrays with Optimized Excitation Weighting for Large Planar Array Applications

The existing subarray optimization approaches find the optimal subarray shapes among all the possible subarray partitions; thus, they are relatively slow and complex due to a huge number of the examined combinations. In this paper, we propose a new two clustered subarrays partitioning schemes for performance optimization of the large planar antenna arrays. The proposed clustered subarrays are capable of simplifying the feeding network, eliminating the grating lobes, minimizing the sidelobe peaks, and improving the taper efficiency. The large planar array elements are first divided into either smaller irregular upladder-like clusters or just like ascending square-rings clusters. Then, their excitation amplitudes are weighted and optimized at the subarray clustered level instead of their original elements level. The novelty of the proposed clusters is its capability to efficiently approximate the actual element excitation amplitudes of the non-uniformly excited large arrays where their amplitudes are gradually increasing from the perimeter to the center of the array which they can be well approximated by ascending square rings around the array center. Moreover, the problem of the grating lobes and the complexity of the array feeding network are managed with these elegant clustered shapes. The genetic optimization algorithm was utilized to optimize the clustered weights such that they best approximate the actual elemental amplitude excitations envelope. Thus, the array taper efficiency does not affected, while minimizing the sidelobe peaks in the clustered subarray patterns. Simulation results demonstrate the efficiency and the superiority of the proposed clustered subarrays over the standard fullarray and the existing regular and irregular subarray methods.