Iterative Algorithm for Indoor Multiple Target Tracking in 3D Space Using mmWave FMCW MIMO Radar

In this study, we address the challenge of matching separately identified targets in the range-azimuth and range-elevation planes, particularly in scenarios involving multiple targets located in the same range. Range-azimuth and range-elevation maps were used in the radar imagery to represent the spatial positions of the detected objects. A range-azimuth map displays the distance (range) and horizontal direction (azimuth) of the targets, providing a top-down view of the environment. A range-elevation map shows the distance and vertical angle (elevation) of the targets, offering a side view to determine how high or low the objects are relative to the radar. We propose a solution that enables 3D positioning by employing an algorithm designed to recursively extract the range, azimuth, and elevation information from a radar data cube built with a nonlinear virtual antenna array. This iterative process was systematically applied across all the frames of the radar data cube to capture the temporal dimensions and perform real-time indoor tracking. Through comprehensive empirical measurements, we demonstrate the efficacy of our approach, achieving precise tracking in scenarios involving a single target and two targets in a permanent back-and-forth movement.