Walking artifacts mitigation for improved heading estimation in a reduced multi-sensor configuration

Global Positioning System (GPS), is predominantly the only source of positioning information capable of working in all kinds of environment throughout the day. The major drawback to this widely accepted positioning technology is its inability to work in indoor environments. Alternatively, various infrastructure based technologies exist such as Wi-Fi, NFC, floor map to assist indoor navigation. However, there still exists a huge gap in the acceptance of such technologies for positioning in GPS-denied environments as additional infrastructure needs to be deployed. Thus the recent research focus is geared towards developing an infrastructure-free system utilizing samrtphone in-built sensors. The research in this paper proposes a novel algorithm capable of mitigating the heading error introduced due to the individuals walking behavior. The algorithm discriminates the pattern in the sensor measurements from the walking against the users actual directional motion. Given this ability the effect of integration error in heading estimation is considerably reduced thereby delivering reliable heading and consequently positioning information. Various test trajectories such as circular, curved, rectangular shaped were considered in an indoor real time environments corresponding to different individuals walking pattern. Test results demonstrated the proposed algorithm effectiveness in bridging the GPS gaps or to aid navigation in an indoor environment.