Metasight: High-Resolution NLoS Radar with Efficient Metasurface Encoding
Timothy Woodford, Kun Qian, Xinyu Zhang
Abstract
A large number of traffic collisions occur as a result of non-line-of-sight (NLoS) obstructions. Recent work has explored NLoS automotive radar sensing systems to detect objects in occluded regions. However, current NLoS radars require substantial ambient reflectors, whose size needs to scale with the desired angular resolution and coverage, impeding their deployment in real-world scenarios. In this paper, we propose Metasight, which leverages carefully designed passive millimeter-wave metasurface reflectors and a novel angular encoding scheme to dramatically reduce the reflector size. The Metasight metasurfaces are fully passive, low cost, and can be fabricated by simply using a 3D printer and copper tape. By processing the reflected signals with a robust angle decoding algorithm on the radar, Metasight achieves high NLoS sensing resolution and wide coverage, with an asymptotically higher space-efficiency than conventional natural or artificial reflectors.