Intelligent metasurface system for automatic near- and far-field tracking of multi-target based on computer vision

Target tracking in wireless communication will become increasingly important in the future. Due to its effective and adaptable control over electromagnetic (EM) waves, it is cheaper, less complex and smaller than traditional antenna manufacturing, a smart and flexible solutions can be provided through digital programmable metasurface (DPM). In this paper, we propose an intelligent metasurface communication system for multi-target tracking, which uses computer vision assistance in artificial intelligence to detect moving targets and position information, and then feeds back to DPM for flexible EM regulation to achieve beam tracking of specific targets. Experimental demonstrations are carried out, including far-field measurement, multi-target tracking and near-field tracking. The proposed method provides a path forward for radio environment tracking, self-adaptive systems, and intelligent wireless communications.