Accurate Indoor Localization with Ultra-Wideband using Spatial Models and Collaboration
Ultra-wideband (UWB) localization is a recent technology that performs competitively with many indoor localization methods currently available. Despite its desirable traits, such as potential high accuracy and high material penetrability, the resolution of non-line-of-sight signals remains a very hard problem and has a significant impact on the localization performance. In this work, we address the peculiarities of UWB error behavior by building models that capture the spatiality as well as the multimodal statistics of the error behavior. Our framework utilizes tessellated maps that associate probabilistic error models to localities in space. In addition to our UWB localization strategy (which provides absolute position estimates), we investigate the effects of collaboration in the form of relative positioning. To this end, we develop a relative range and bearing model and, together with the UWB model, present a unified localization technique based on a particle filter framework. We test our approach experimentally on a group of 10 mobile robots equipped with UWB emitters and extension modules providing inter-robot relative range and bearing measurements. Our experimental insights highlight the benefits of collaboration, which are consistent over numerous experimental scenarios. Also, we show the relevance, in terms of positioning accuracy, of our multimodal UWB measurement model by performing systematic comparisons with two alternative measurement models. Our final results show median localization errors below 10 cm in cluttered environments, using a modest set of 50 particles in our filter.