Precision scalable indoor localization using ultra-wideband impulse radio and crowdsourcing

Abstract

This dissertation is concentrated on positioning and tracking algorithms to enable scalable precision indoor localization using ultra-wideband impulse radio (UWB-IR) by exploiting crowdsourcing. As indoor positioning systems require high precision, UWB technology stands out by offering low-cost and highly accurate positioning with accuracy reaching within a few centimeters with low latency. However, non-line-of-sight (NLOS) conditions and environmental noise impair the accuracy. In addition to the positioning, the crowdsourcing method is introduced to gather real-time data. The devices collaborate with each other to achieve the localization. In order to achieve localization, the procedure involves collection of crowdsourcing data, NLOS identification, positioning algorithms (least squares method and Broyden-Fletcher-Goldfarb-Shanno method) and tracking algorithms (Kalman filter). In the simulation conducted on python, the localization results are achieved with high accuracy in line-of-sight (LOS) scenario. In NLOS scenario, the estimated positions are less accurate. When the algorithms work together, the results are with best performance.