Performance evaluation of a new Kalman filter based peer‐to‐peer tracking scheme for indoor environment

Abstract

Peer‐to‐peer tracking through smartphone sensor data is in demand due to its usefulness in location‐based services. A person carrying a smartphone device could be tracked by another smartphone through real time signal processing. Due to the distortion of GPS signals in indoor environment, Kalman filter based data fusion techniques are popularly applied to integrate various sensor data. Such an approach suffers failure in the absence of external aiding and thus entails peer tracking only through the smartphone's navigation sensor data. In this context, accurate estimation of heading error between the leaders and followers' trajectory is very much crucial. The present work demonstrates one novel Kalman filter‐based measurement matching approach for accurate estimation of the aforesaid heading error. Less than 1 meter of accuracy in the final position estimation has been achieved through this method which is comparable with other state of the art techniques as reported in literatures. Moreover, the system does not depend on any external aiding which makes it adaptable to any unknown indoor location.