3-Axis magnetic field mapping and fusion for indoor localization

As location-based services have grown increasingly popular, they have become limited by the inability to acquire accurate location information in indoor environments, where the Global Positioning System does not function. In this field, magnetometers have primarily been used as compasses. As such, they are seen as unreliable sensors when in presence of magnetic field disturbances, which are frequent in indoor environment. This work presents a method to account for and extract useful information from those disturbances. This method leads to improved localization in an indoor environment. Local magnetic disturbances carry enough information to localize without the help of other sensors. We describe an algorithm allowing to do so as long as we have access to a map of those disturbances. We then expose a fast mapping technique to produce such maps and we apply this technique to show the stability of the magnetic disturbances in time. Finally, the proposed localization algorithm is tested in a realistic situation, showing high-quality localization capability.