Device-Free Tracking via Joint Velocity and AOA Estimation With Commodity WiFi

Abstract

Device-free localization and tracking (DFT) is a novel technique which can estimate the location of a target without equipping it with any devices. Existing DFT systems mainly rely on machine learning techniques with a labor-intensive training. Recently some training-free DFT systems are designed by deriving the angle-of-arrival (AOA) using the dedicated hardware. However, limited by the hardware imperfection, realizing a training-free DFT system using the commodity Wi-Fi is still a challenging task to solve. To address this issue, we develop a novel DFT system which can track the target motion via both target velocity and AOA estimations simultaneously. First, the motion-induced phase shifts are refined from extremely noisy channel state information (CSI) measurements to detect the Doppler shift based on the signal superposition analysis. Then, according to the phased-array signal processing, we realize joint Doppler velocity and AOA estimation of the target path under the compressive sensing framework. It formulates joint Doppler velocity and AOA estimation as a two-dimensional sparse reconstruction problem, which can achieve a high accuracy to further estimate the target velocity and track the target motion at a decimeter level. We implement the proposed DFT system on commercial Wi-Fi devices and validate its performance with extensive evaluations in three indoor scenarios.