A Robust Indoor Pedestrian Localization Approach Against Human Body Shadowing for UWB-Enabled Smartphones

Embedding the ultra-wideband (UWB) chips inside smartphones makes UWB an attractive and promising technology for centimeter-level indoor pedestrian tracking. However, the daily life wearing diversity of the UWB-enabled smartphones may lead to meter-level localization errors. This happens due to the human body shadowing (HBS) between the line-of-sight channel of a smartphone (tag) and infrastructure sensors (anchors). Existing solutions have countered this challenge by detecting and mitigating the degree of HBS using extra perceptions (signal quality and walking direction) specific to tag’s on-body placement. This article proposes a robust indoor pedestrian localization approach for UWB-enabled smartphones against HBS independent of any extra perception. In this regard, a UWB anchor-pairs positioning fusion (UAPF) approach is proposed in three steps. First, the bilateration concept is adopted for UWB pedestrian localization. Second, a new term, named neighbor distance qualifier, is introduced to qualify the valid UWB anchor-pairs-based pedestrian localization. Third, a linear Kalman filter (KF) is modeled to smooth the pedestrian’s state. The experimental comparative results validated that UAPF outperformed the standard extended KF by 42%, 65%, 62%, and 58% and the Qorvo location engine by 4%, 50%, 40%, and 55%, respectively, for daily life smartphone-carrying contexts (texting, calling, swinging, and pocketing).