A Kalman filter for UWB positioning in LOS/NLOS scenarios

Compared with other indoor localization technology, Ultra-wide Band (UWB) has the advantages of high precision, strong communication ability, wide coverage and convenient deployment, etc. As a result, UWB is a very popular solution for most indoor positioning applications. However, the performance of UWB signal is susceptible to several factors, among which Non-Line-of-Sight (NLOS) propagation is the most common and also a major source of measurement errors. Among the existing localization algorithms, Kalman filtering (KF) has received wide attention as it is a classical method for state estimation of dynamic systems. In this paper, a Distance-KF method and a Position-KF method are designed based on the environment characteristics of indoor LOS/NLOS scenarios. The two strategies are studied on the WxRobot UWB platform. The results verify that the two methods can improve the accuracy and the best improvement of the average error is 27.4% in the Position-KF method of LOS scenario.