Development of UWB NLOS Model Aided by IMU in Indoor Positioning Systems

Indoor positioning systems need to meet both functional, quality requirements, and development constraints. Functional requirements include the capabilities of positioning, navigation, and timing. Quality requirements attributes and development constraints include accuracy, robustness, reliability, cost-effectiveness etc. The sensor technologies provide the physical functions to meet the functional requirements. The sensors working in some conditions would cause positioning errors and reduce quality. In this research, we explore the Ultra-Wideband (UWB) sensor error modelling approach to achieve a higher quality level of accuracy and reliability and reduce the cost of hardware deployment. UWB technology provides high accuracy in positioning but suffers from non-line-of-sight (NLOS) effects in indoor environments. We use Inertial Measurement Unit (IMU) as an external sensor aiding the NLOS modelling approach. In this paper, we propose a segmentation error model based on the Relative Yaw Angle (RYA), through which the combination of this method and the weighted least squares method can lead to an improvement of more than 60% in the positioning accuracy of UWB systems in the NLOS environment This approach is validated using experimental results.