Indoor mobile localization system using UKF and pre-filtering

In this paper, we present the practical application of unscented Kalman filter (UKF) for Indoor mobile localization system using ultrasonic sensor. It is true that many kinds of localization techniques have been researched for several years in order to contribute to realize ubiquitous system; particularly, such a ubiquitous system needs much highly precise accuracy to be practical and efficient. However, a number of localization systems for indoor space do not unfortunately have enough accuracy to establish any special task such as precise position control of moving target even though they have comparatively high developed cost. Therefore, we developed indoor mobile localization system having high localization performance; specifically, the unscented Kalman filter is applied for improving the localization accuracy. In addition, we also present the additive filter named ´pre-filtering´ to compensate the performance of estimation algorithm. Pre-filtering has been developed to overcome bad effect from unexpected external noise so that localization through unscented Kalman filter has come to be stable. Moreover, we tried to demonstrate the performance comparison of the unscented Kalman filter and another estimation algorithm, such as the unscented particle filter (UPF) through simulation for our system.