Variable-Dimension Differential Evolution for Ring-Count-Optimized Sparse Concentric Ring Array Synthesis

Abstract

This communication proposes a variable-dimension differential evolution (VDDE) algorithm for synthesizing uniform amplitude sparse concentric ring arrays (CRAs) with minimized peak sidelobe level (PSL). Different from conventional fixed-ring-count methods, the VDDE simultaneously optimizes the ring count, radii, and element counts per ring under the constraints of aperture size, minimum inter-element spacing, and total element budget. A mapping-based over-provisioned encoding scheme is proposed to handle the variable-dimension individuals, and targeted mutation, crossover, and selection operators are also introduced. Numerical results show that the proposed VDDE achieves lower PSL with even fewer elements than state-of-the-art algorithms, validating its effectiveness and superiority.