Aiming at shortcomings of low localization precision of pedestrian dead reckoning（ PDR）

a PDR algorithm based on particle filtering（ PF） is proposed. The positioning algorithm uses the built-in accelerometer to detect and count the steps

and gyroscope and magnetometer are used for direction estimation. According to information of map matching

PF algorithm is used for data filtering fusion. The experimental results show that the PDR positioning algorithm based on PF improves positioning precision.