Synthesis of nonuniform linear antenna arrays based on subarray division

A novel hybrid optimization algorithm is proposed and applied to the subarray partitioning of linear non-uniform antenna arrays. The positions and excitations of the array elements are optimized using Invasive Weed Optimization (IWO) and Convex Optimization (CO), respectively. The proposed hybrid optimization algorithm can combine the capability of IWO to solve complex optimization problems and the superior global convergence properties of CO. To further reduce the design complexity and cost, the K-means clustering algorithm is applied to fulfill the subarray partitioning of the antenna array. A newly selection method of the initial cluster centers is introduced. To evaluate the effectiveness and superiority of the proposed algorithm, radiation patterns are synthesized under both contiguous and non-contiguous subarray partitioning conditions. Compared with relevant optimization approaches, the proposed hybrid optimization algorithm can effectively suppresses the peak sidelobe level (PSLL).