Dimension-Reduced Optimization for Uniform Near-Field Synthesis of Irregular Arrays

A novel optimization method of a significantly reduced optimizing dimension is presented to quickly synthesize the uniform near fields of irregular arrays. In this method, the element excitations of the array are no longer executed as the direct optimization parameters. Instead, we build a new search space made of the smooth reference masks (SRMs) and perform the optimization directly based on the parameters of the SRMs rather than the element excitations. To obtain the desired uniform near fields, only two critical parameters of the SRMs need to be optimized. Comparing the direct element-excitation optimizations, the proposed method significantly reduces the optimizing dimension and makes the synthesis much more efficient and faster. Several numerical examples and their comparisons with the well-developed convex optimization algorithm were presented. The results demonstrate the effectiveness and the significant reduction of time consumption of the proposed method.