Magneto-Inertial Dead-Reckoning Navigation with Walk Dynamic Model in Indoor Environment

We tackle the problem of pedestrian indoor navigation with Magneto-Inertial Dead Reckoning technology (MIDR) with the integration of the data provided by a Magneto-Inertial Measurement Unit (MIMU). This method is well-known in the literature and very efficient if the spatial distribution of the magnetic field is nonuniform and the gradient of sufficiently high magnitude. However, the quality of the magnetic information decreases as the pedestrian moves toward a weak magnetic zone. We propose, characterize and test a new correction technique of the MDIR velocity based on the detection of walking steps and dynamical modeling of the walk itself.