Impact of the unit cell distribution of 1-D dynamic metasurface antennas on the performance of a computational imaging system

Computational imaging (CI) enables to bypass the significantly demanding sampling requirements imposed by synthetic aperture radar (SAR) techniques. For that purpose, CI systems rely on compressive antennas to achieve a physical-layer compression of the scene information. In this context, among the different compressive antenna implementations, dynamic meta-surface antennas (DMAs) are of significant interest due to their compactness, ease of fabrication, and their ability to reconstruct the scene information without the need of large bandwidths. This contribution analyzes the impact of using different unit cell distributions for a 1-D DMA on the performance of a single-pixel CI system. Results show that the use of a sparse unit cell distribution does not jeopardize the imaging performance of the system, and, in addition, it enables to reduce the hardware complexity of the antenna.