Magnetic field navigation in an indoor environment

Abstract

This paper describes a method that has been developed to aid an inertial navigation system when GNSS signals are not available, by taking advantage of the uniqueness of magnetic field variations. Most indoor environments have many different features (ferrous structural materials or contents, electrical currents, etc.) which perturb the Earths natural magnetic field. The variations in the magnetic field in indoor environments can be used as a way to identify a users position, and possibly orientation, because the 3-dimensional magnetic field varies significantly as a function of position. Using relatively inexpensive 3-axis magnetic field sensors, it is possible to estimate a users location in an indoor environment. This paper presents algorithms and results from two different scenarios, based on a real data map-matching scenario, and a leader-follower scenario. In the map matching scenario, it is assumed that a map of the magnetic field in the indoor environment has been previously collected. Magnetic field measurements are then compared to the previously generated map to get an absolute position. Results using real data showed very good (decimeter-level) positioning performance. For the leader-follower scenario, the lead vehicle moves through the environment, measures the magnetic field, and sends information about what it has observed to the follower vehicle, which attempts to follow the same path as the lead vehicle. Results show that it is possible for a follower vehicle to follow a lead vehicle using this approach, especially if additional constraints (such as wall boundaries) can be applied to further constrain the trajectory. Finally, this paper describes some practical calibration issues that must be considered when implementing a magnetic field navigation system.