NLOS Error Mitigation in a Location Estimation of Object based on RTLS Using Kalman Filter

Abstract

The real time locating systems (RTLS) based on a set of location access points (LAP) receiving radio frequency (RF) and associated computing equipment are models which determine the position of a transmitting device relative to the placement of the aforementioned receivers that is capable of reporting that position within several minutes of the transmission used for determining the position of the transmission. The RTLS has been widely applied in various areas such as parts replenishment, vehicle management, yard management, container management and assert management. The most challenging issues which render to reach the required accuracy for the time-based location system are non line of sight (NLOS) propagation problems. The NLOS environment prevents a LAP from receiving RF signals from an RTLS tag. Namely, RTLS which tracks location by using time difference of arrival (TDOA) cannot exactly receive RF signals from an RTLS tag attached to the object. Therefore, the estimated location of an RTLS tag is different from a tag location. Accordingly, the Kalman filtering techniques are applied to RTLS to improve its accuracy. The Kalman filter is rather a recursive estimator than a batch estimator. Because, it gradually converges estimation by using a set of the estimated data one by one. We propose an location filter designed using the Kalman filter for location mitigation in an NLOS environment. We used the linear equation and the observation equation to apply this system to the Kalman filter