Device-Free Pedestrian Tracking Using Low-Cost Ultrawideband Devices

Abstract

Ultrawideband (UWB) attracts extensive attention for the Internet-of-Things applications, especially the fine-grained location-based services. Rather than active tracking, this article explores the UWB-based device-free pedestrian tracking problem. Concerning the challenges of fine-grained passive tracking for the low-cost commodity UWB devices, we propose a variance-based temporary-spatial (VATS) mapping algorithm, which relieves the background interference from the perspective of the Bayesian framework. Moreover, a particle filter algorithm has been designed to track the position likelihood changing and to avoid likelihood ambiguity. Experimental results show that the proposed VATS mapping algorithm achieves 50th and 90th percentile errors 0.156 and 0.272 m, respectively, which is promising for practical applications.