Synthesis of Subarrayed Large Linear Arrays by a Hybrid Genetic Algorithm Integrated with Convex Programming.

In this paper, we devise a new hybrid method for pattern synthesis of large linear arrays. To simplify the beamforming network, the uniform linear array is divided into multiple non-overlapped subarrays with contiguous elements and is only weighted at subarrays. Unlike previous approaches, the proposed method introduces a special vector to specify the subarray partition and converts the subarray pattern synthesis to the problem of finding the optimal subarray partition and subarray weighting vectors simultaneously. To solve this problem efficiently, the genetic algorithm is utilized to find the optimal partition scheme from all possible subarray partitions. For the fixed subarray partition, the optimal subarray weighting vector is obtained by solving the convex sub-problem. Numerical results demonstrate the superiority of our method over the state-of-the-art algorithm. [ABSTRACT FROM AUTHOR]