Real-time PDR based on resource-constrained embedded platform

Abstract

Standalone inertial navigation system (INS) in indoor pedestrian positioning is becoming imminent as the researchers exploit its small form factor and low power requirement. This will result in small-size, low-power wearable devices that are not obtrusive to the users and yet provide sufficiently accurate pedestrian localization and tracking within. At this stage, most of the recent INS-based indoor pedestrian positioning systems still have to interface with other computing machines such as a laptop or smartphone to perform computationally demanding algorithms. Most of the existing techniques operate in off-line and not real-time mode. In this paper, we propose a real-time indoor pedestrian dead-reckoning system based on embedded INS. The results show that our system successfully track the distance travelled by pedestrians up to an error of three percent with a position update interval less than a second.