Particle filter for context sensitive indoor pedestrian navigation

Novel particle filter design combines foot mounted inertial (IMU), bluetooth low energy (BLE) and ultra-wideband (UWB) technologies along with map matching into a seamless integrated navigation system for indoors. The system was evaluated by 10 test walks along an 80m long indoor track including stairs and running. 95% of the time the average error for a particle was below 3.1 m with filter completion success rate of 90%. Furthermore, a system without UWB using only IMU, BLE and map matching achieved an average error for a particle to be below 3.6 m with filter completion success rate of 70%. The selected technologies and sensors are affordable and easily deployable. Inertial measurement unit's characteristics complement the disadvantages in the rf technologies and vice versa. The code for the rover can be implemented on a modern mobile device together with a foot mounted IMU