Neuro‐Driven Mechanically Programmable Metasurface for Electromagnetic Environment‐Illusion Generation
Year: 2026
Type: article
Abstract: ABSTRACT Invisibility cloaks are evolving beyond passive scattering suppression toward the active control of radar perception. However, on‐demand generation of electromagnetic environment illusions remains challenging. Passive metasurfaces are fixed after fabrication, whereas electrically reconfigurable metasurfaces usually require dense bias networks and continuous electrical holding power, which complicate large‐aperture integration and introduce steady‐state energy consumption. Here, we report an X‐band neuro‐driven mechanically programmable metasurface (NMPM) that generates prescribed electromagnetic signatures through mechanically retained coding states. To map target scattering profiles onto implementable mechanical configurations, we develop a calibration‐assisted, physics‐guided neural programming framework. This framework compensates for coupling‐induced model mismatch and addresses the non‐unique inverse mapping from desired electromagnetic responses to discrete array states. The generated states are physically written by rotating reflective meta‐atoms optimized for linearly polarized illumination, allowing the programmed configuration to be retained without continuous electrical bias. Microwave experiments on two representative target environments demonstrate that the NMPM reproduces the dominant main‐lobe directions and overall sidelobe envelopes of the target scattering profiles. Quantitative measurements yield cosine similarities of 94%–98%, with consistent agreement under the tested incidence‐angle and frequency variations. These results demonstrate a mechanically retained and algorithmically calibrated route toward adaptive radar deception, electromagnetic scene engineering, and reconfigurable sensing architectures.
Source: Laser & Photonics Review
DOI: 10.1002/lpor.72123
Language: English
Subfield: Aerospace Engineering
Field: Engineering
Domain: Physical Sciences
Open Access status: bronze
OpenAlex ID: W7223465815