Surrogate-Assisted Design of Checkerboard Metasurface for Broadband Radar Cross-Section Reduction

Abstract

Metasurfaces have been extensively exploited in stealth applications to reduce radar cross section (RCS). They rely on the manipulation of backward scattering of electromagnetic (EM) waves into various oblique angles. However, arbitrary control of the scattering properties poses a significant challenge as a design task. Yet it is a principal requirement for making RCS reduction possible. This article introduces a surrogate-based approach for rapid design optimization of checkerboard metasurfaces. Our methodology involves fast metamodels, and a combination of surrogate-assisted global optimization with local, gradient-based tuning. It permits an efficient control of the EM wave reflection characteristics, and ensures arriving at that the globally optimum solution within the assumed parameter space. The design procedure is fully automated. The framework is employed to develop a novel broadband checkerboard metasurface, where the RCS reduction is fundamentally based on the backward scattering manipulation carefully controlled by simultaneous adjustment of the unit cell dimensions. The properties of the structure are demonstrated using simulated monostatic and bistatic RCSs. The proposed metasurface exhibits 6 dB RCS reduction within the frequency range from 16 to 37 GHz. The numerical results are validated using physical measurements of the fabricated prototype. Experimental data indicates that the relative RCS reduction bandwidth is 83 percent, which makes the proposed structure outperforming the designs reported in the literature.