Ultrawideband Ranging in Dynamic Dense Human Networks

Non-line-of-sight (NLoS) conditions are a known performance limiting factor in ultrawideband (UWB) ranging accuracy, particularly in environments with densely packed dynamically moving humans, i.e., crowds. As UWB technology is recently seeing wide deployment in consumer devices, the effects of NLoS conditions due to human body shadowing will greatly degrade the performance of related ranging and localization applications. We thus propose the round-robin ranging protocol, an extension of alternative double-sided two-way ranging, which scales well for high rate sampling of dense networks and further propose a cooperative statistical approach based on multidimensional scaling (MDS) to mitigate NLoS conditions. Experimental validation was performed in a densely packed dynamically moving human environment. Nine subjects walked naturally within a 16-m
²
motion capture area while wearing UWB sensors at the foot, wrist, torso, and head and reflective optical motion capture markers. UWB accuracy of the 36 intersubject ranges was computed as the difference in range estimates between the UWB sensors and the optical motion capture system. Results showed that the proposed MDS approach reduced ranging accuracy errors by 11%, 13%, 16%, and 32% at the foot, wrist, torso, and head, respectively. Body placement had a significant effect on the ranging performance in that transceivers placed on the head and foot reduced ranging errors by 84% and 63% compared with the traditional torso placement. These results suggest that the proposed approaches could significantly improve the performance in UWB ranging applications involving human bodies such as human movement monitoring, athletics, logistics, and social contact tracing applications.