Metasurface Design Paradigm Inspired by Connected Component Analysis to Realize Broadband Surface Wave Manipulation

This work presents a broadband surface wave (SW) radiation control design based on gradient metasurfaces (GMs). To achieve stable broadband beams and efficient radiation, GMs must satisfy two requirements simultaneously: a flat phase response, and operation within the SW bandgap to suppress edge scattering. To address these needs, this work develops a broadband GM design paradigm inspired by connected component analysis (CCA). This paradigm employs a nonconvex optimization strategy for the iterative optimization of pixel-level topologies; by incorporating prior GM topology information, it can constrain solutions within a predefined range, thereby enabling optimization with low data volume. For the GM bandgap design, this paradigm integrates the design theory of uniplanar compact electromagnetic bandgaps, leveraging the large equivalent inductance from long topological contours to reduce the bandgap frequency. Innovatively, it incorporates CCA algorithms from the field of image processing into the design, using CCA to extract the topological perimeter as a factor for regulating fitness. This avoids time-consuming eigenmode solving, significantly enhancing design efficiency. To validate the paradigm, two radiation prototypes (a pencil-beam prototype and a flat-top-beam prototype) were designed. Under the scattering mode, beamforming is manipulated based on Huygens' first principle, and the measured results show high consistency with the simulated results. In which, the pencil-beam antenna achieves an aperture efficiency of 54% and a 3-dB gain bandwidth of 25%, while the flat-top-beam antenna maintains stable flat-top beam characteristics in the 9-11 GHz frequency range. Combining excellent broadband performance and high design efficiency, this design provides a novel technical pathway for SW radiation control.