A pedestrian tracking system using group mobility information

While tracking in outdoor environments is easily achieved by GPS, indoor tracking, especially in an underground environments is not possible. One common technique for indoor tracking is to employ inertial sensors, but the main problem with inertial pedestrian tracking systems is that they accumulate errors and require periodic recalibration to avoid errors. In this paper, we present a pedestrian tracking system using group mobility information that several pedestrians form stable moving clusters. Group mobility information is shared and cooperated in overcoming the accumulated tracking error inherent the Inertial Measurement Unit (IMU). A prototype is implemented in smartphone and the experiment results show that the location error is reduced up to 43%, compared to traditional dead-reckoning tracking methods.