Key points are not available for this paper at this time.
Magnetic field matching (MFM) positioning is one of the mainstream methods of consumer indoor positioning, which has attracted great attention from academia and industry. However, efficient methods for constructing magnetic field maps and matching positioning methods with high stability still need to be developed. This study proposes an indoor MFM positioning scheme enhanced by consumer-grade inertial measurement units that can efficiently generate a magnetic field grid map and achieve robust matching positioning, without the need to actively calibrate the magnetometer bias. When generating the magnetic field map, the proposed method employs a pedestrian positioning and orientation system to efficiently collect data by releasing individual behavioral constraints and reducing the number of control points required. Moreover, a magnetometer bias autocalibration method, aided by the precise attitude, is proposed to simplify the data collection process. During real-time positioning, the position and attitude generated by pedestrian dead reckoning (PDR) are used to generate the differential magnetic field strength in the sensor frame; this achieves matching positioning that is independent of the magnetometer bias. Furthermore, an estimation method for the MFM position noise, based on the magnetic field gradient, is proposed to improve the accuracy of MFM/PDR integrated positioning. Several experiments were conducted to verify the feasibility and performance of the proposed scheme. Only slight differences were found between the magnetic field maps using three different smartphones, showing that the proposed scheme can efficiently generate a high-precision magnetic map. The positioning results of multiple tests conducted using eight different smartphones revealed that the proposed scheme achieves continuous and robust meter-level positioning accuracy.
Kuang et al. (Mon,) studied this question.