Programmable Metasurfaces for Computational DoA Estimation: Experimental Validation

We present the application of programmable coded-metasurfaces for computational imaging (CI)-driven characterization of wireless channels at microwave frequencies. The presented CI system consists of a dynamic metasurface antenna (DMA) that can encode the channel information through its spatio-temporally incoherent, quasi-random transfer function, compressing the measurements into a single port. The quasi-random radiation patterns of the DMA are generated by means of binary-modulating the metamaterial elements forming the DMA aperture using p-i-n diodes. Different from conventional phased-array-based sensing architectures relying on raster scans, the proposed DMA-based CI-driven channel characterization approach eliminates the raster-scan requirement, replacing it with quasi-random measurements, while also eliminating the need to use complex phase shifters. We experimentally synthesize the DMA-CI system, and demonstrate its use for direction-of-arrival (DoA) estimation.