RF-Based Initialisation for Inertial Pedestrian Tracking

Location information is an important source of context for ubiquitous computing systems. We have previously developed a wearable location system that combines a foot-mounted inertial unit, a detailed building model and a particle filter to locate and track humans in indoor environments. In this paper we present an algorithm in which a map of radio beacon signal strengths is used to solve two of the major problems with the original system: scalability to large environments and uncertainty due to environmental symmetry. We show that the algorithm allows the deployment of the system in arbitrarily large buildings, and that uncertainty due to environmental symmetry is reduced. This reduction allows a user to be located after taking an average of 38 steps in a 8725 m2 three-storey building, compared with 76 steps in the original system. Finally, we show that radio maps such as those required by the algorithm can be generated quickly and automatically using the wearable location system itse...

• Type: Conference
• Year: 2009
• Where: PERVASIVE
• Authors: Oliver Woodman, Robert Harle
• Keywords: Applied Computing, Environmental Symmetry, Particle filters, PERVASIVE 2009, Wearable Location