Optimal data fusion for pedestrian navigation based on UWB and MEMS

Indoor pedestrian navigation is probably a very challenging research area. In this context, an optimal data fusion filter that hybridises a large set of observations: angles of arrival (AOA), time differences of arrival (TDOA), accelerations, angular velocities and magnetic field measurements is presented. The coupling of UWB and MEMS data relies on an extended Kalman filter complemented with specific procedures. Geometry based algorithms and a RANSAC paradigm that mitigates the non line of sight (NLOS) UWB propagation are detailed. The benefit of the solution is evaluated and compared with the pure inertial positioning system.