Contact-Free Pedestrian Tracking Using Massive MIMO-OFDM Communication System

This paper establishes an indoor distributed radar-like prototype using a sub-6 GHz massive multiple-input multiple-output (MIMO) communication system. Based on the prototype, the contact-free human tracking performance is experimentally evaluated. Instead of extracting the location related geometrical metrics (angle and distance), we propose to localize the pedestrian through the channel state information (CSI) directly. Inspired by the concept of synthetic aperture, a particle filter-based hologram tracking algorithm is proposed. The experimental results show that the proposed algorithm can achieve 20.1-cm mean tracking accuracy in real-time.