In order to solve accumulated error problem of low cost micro-electro-mechanical inertial navigation system

a new real-time indoor pedestrian navigation system based on fusion of pedestrian dead reckoning（ PDR）and ultra wideband（ UWB） wireless positioning is proposed. The accelerometer and magnetometer are used to calculate the initial attitude; inertial navigation algorithm is deduced by considering the filtering error estimation;the"and"logic is used in step detection algorithm by relying on the accelerometer and gyroscope; zero velocity update（ ZUPT） is carried out to provide velocity error measurements and UWB system is used to provide position error measurements for the filter; an extended Kalman filter with the outliers identification mechanism is designed for data fusion. The proposed pedestrian navigation algorithm is verified

and the proposed algorithm can effectively improve the positioning position of pedestrian compared with the traditional method. In the experiment

the pedestrian navigation algorithm can obtain the positioning error which is less than 0. 2 m. The proposed navigation system is stable and divergence problem does not exist.