Multifunctional Intelligent Metamaterial Computing System: Independent Parallel Analog Signal Processing

Abstract

Analog computing based on miniaturized surfaces has gained attention for its high-speed and low-power mathematical operations. Building on recent advances, an anisotropic space-time digital metasurface for parallel and programmable wave-based mathematical operations is proposed. Using frequency conversions, our metasurface performs 1st-order and 2nd-order spatial differentiations, integrodifferential equations, and sharp edge detection in spatially encoded images. The anisotropic nature of the meta-particle enables independent and simultaneous operations for two orthogonal polarizations. Reconfigurability is achieved through tunable gate biasing of an indium tin oxide layer. Illustrative examples demonstrate that the metasurface's output signals and transfer functions closely match ideal transfer functions, confirming its versatility and effectiveness. Unlike other wave-based signal processors, the design handles wide spatial frequency bandwidths, even with high spatial frequency inputs.