A Low-cost Pedestrian Navigation System Including In-flight Parameter Correction

Y. Kim, A. Cho, O. Kim, C. Kim, C. Kee

Abstract:	This paper presents a GPS/PDR (Pedestrian Dead Reckoning) algorithm for pedestrian navigation including in-flight step length parameter correction. GPS compensates the position error from PDR and estimates the step length parameter. Kalman filter is used as an estimating technique for integrating the stride length and GPS measurement. The heading is calculated using GPS heading and gyro output. According to pedestrian location based service takes the limelight as a new growth engine, need for the pedestrian navigation is increasing. Among the existing navigation system, GPS can be easily applied to the pedestrian navigation by using low cost equipment. But in urban environments, serious position error occurs when GPS signal is blocked by high building or tree. In order to solve this blockage problem, one of the solutions is integration of GPS and other navigation system. And GPS/INS is the famous integration algorithm. In car navigation and aircraft application, GPS/INS is well performed. But, in pedestrian navigation application, the classical GPS/INS approach used for vehicle navigation is not suitable[1]. The main problem to solve is the alignment of the IMU (Inertial Measurement Unit). An alternative is to use accelerometer signal pattern rather than its value to count the steps, this is PDR. The PDR is a kind of navigation system that provides pedestrian’s position using own inertial sensors. The main algorithm of PDR is composed of step detection, step length estimation, heading estimation algorithms. PDR has a position shift error and slow propagation characteristics. The performance of PDR depends on the accuracy of step length estimation algorithms. But the step length varies with health condition, ground status and individual gait characteristics, etc. Therefore step length correction is needed for improving the position accuracy of PDR. In this paper, GPS/PDR algorithm for pedestrian navigation including in-flight step length parameter correction is presented. The step length is modeled as a linear combination of constants and step frequency and step length parameter is corrected by using GPS position. Kalman filter is used as an estimating technique for step length parameter correction. The corrected parameters are used to improve the position accuracy. Low-cost GPS receiver and MEMS IMU are used. The sensor module is mounted on the top of the head. The performance of the proposed algorithm is verified through the actual walking tests.
Published in:	Proceedings of the 2011 International Technical Meeting of The Institute of Navigation January 24 - 26, 2011 Catamaran Resort Hotel San Diego, CA
Pages:	908 - 915
Cite this article:	Kim, Y., Cho, A., Kim, O., Kim, C., Kee, C., "A Low-cost Pedestrian Navigation System Including In-flight Parameter Correction," Proceedings of the 2011 International Technical Meeting of The Institute of Navigation, San Diego, CA, January 2011, pp. 908-915.
Full Paper:	ION Members/Non-Members: 1 Download Credit Sign In