Cost-efficient UWB and PDR Hybrid Pedestrian Positioning System Performing Incremental Position Estimation

The fusion of Pedestrian Dead Reckoning (PDR) and Radio frequency(RF) technologies with complementary advantages has been broadly applied in modern Indoor Positioning System (IPS). In this paper, a novel Hybrid Pedestrian Positioning System (HPPS) based on two UWB anchors and IMU measurements is proposed to achieve accurate and stable tracking in long distance travelling indoor. Unlike other existing works, HPPS does not require prior knowledge of a initial position to start with. Pedestrian initial position is estimated by First Reliable Step Identification algorithm based on UWB ranging measurements and pedestrian step length. Afterwards, the system iteratively update the position using Incremental Position Estimation algorithm aided by gyroscope. HPPS not only effectively reduces position error caused by sensor bias and step detection error, but also adapts to different attachments (handheld, in the pocket, etc.) and moving modes. As evaluated with four walking experiments, HPPS performs accurate long distance tracking and achieves a maximum of 89% improvement on Average Position Error compared with typical PDR algorithm.