14 September 2020 in IEEE Sensors Journal
Magnetic field localization utilizes position dependent and time persistent distortions of the earth magnetic field. These distortions are introduced by stationary ferromagnetic material in the environment and can be stored in a map to enable localization. Estimating the position of a magnetometer with these distortions requires a calibration of the sensor to enable the matching of the measurements to the map. Typically, the calibration is performed in a prior step and requires specific maneuvers like sensor rotations in a homogenous field. The goal of the maneuvers is to render the calibration parameters observable. For heavy platforms, e.g., cars, trains and driverless transport systems in factories, performing special maneuvers is cumbersome or even impossible. In addition they operate in an environment with an inhomogeneous magnetic field. To address this issue, this paper proposes a novel method that exploits the magnetic field distortions to render the calibration parameters observable. To simplify the calibration process, the calibration parameters are estimated simultaneously with the position of the platform. The method employs a Rao-Blackwellized particle filter that reduces the computational complexity and enables real time processing. The feasibility of the method is shown in an evaluation with measurements of a magnetometer mounted on a model train. The results show a high accuracy of the position and calibration parameter estimation.
23 February 2023 in NAVIGATION: Journal of the Institute of Navigation
Magnetic field localization is based on the fact that the Earth’s magnetic field is distorted in the vicinity of ferromagnetic objects. When ferromagnetic objects are in fixed positions, the distortions are also fixed and, thus, contain location information. In our prior work, we proposed a simultaneous localization and calibration (SLAC) algorithm based on a Rao-Blackwellized particle filter that enables magnetic train localization using only uncalibrated magnetometer measurements. In this paper, a lower-complexity version of the SLAC algorithm is proposed that only estimates a subset of calibration parameters. An evaluation compares the full and reduced SLAC approach to a particle filter in which the magnetometer is pre-calibrated with a fixed set of parameters. The results show a clear advantage for both SLAC approaches and that the SLAC algorithm with a reduced set of calibration parameters achieves the same performance as the one with a full set of parameters.
27 December 2023 in IEEE Transactions on Wireless Communications
Multipath-enhanced device-free localization (MDFL) systems determine presence and location of objects and users not necessarily equipped with localization devices. For localization, MDFL systems exploit user-induced changes in the power of all received signal components, including both line-of-sight and multipath components (MPCs). In this work, we therefore provide a statistical fading model that describes user-induced changes in received power specifically for MPCs. The model is derived and validated empirically using an extensive set of wideband and ultra-wideband measurement data. Since the localization performance of MDFL systems strongly depends on the information about the propagation paths within the wireless network, we further propose a Bayesian calibration approach that estimates the location of the reflection points of MPCs caused by single-bounce reflections. For MPCs caused by single-bounce reflections, the solution space of possible locations of reflection points is constrained to the delay ellipse, which allows the formulation of a computationally efficient one-dimensional estimation problem. Eventually, the problem is solved by sequential Bayesian estimation. The applicability of the proposed approach is demonstrated and evaluated using measurement data. Independent of the underlying measurement system, the Bayesian calibration approach is shown to robustly estimate the locations of the reflection points in different environments. Finally, the localization results of MDFL for an indoor scenario confirm the applicability of the Bayesian calibration approach.
21 November 2022 in IEEE Access
In this paper, we show how to analyze the achievable position accuracy of magnetic localization based on Bayesian Cramér-Rao lower bounds and how to account for deterministic inputs in the bound. The derivation of the bound requires an analytical model, e.g., a map or database, that links the position that is to be estimated to the corresponding magnetic field value. Unfortunately, finding an analytical model from the laws of physics in not feasible due to the complexity of the involved differential equations and the required knowledge about the environment. In this paper, we therefore use a Gaussian process (GP) that approximates the true analytical model based on training data. The GP ensures a smooth, differentiable likelihood and allows a strict Bayesian treatment of the estimation problem. Based on a novel set of measurements recorded in an indoor environment, the bound is evaluated for different sensor heights and is compared to the mean squared error of a particle filter. Furthermore, the bound is calculated for the case when only the magnetic magnitude is used for positioning and the case when the whole vector field is considered. For both cases, the resulting position bound is below 10 cm indicating an high potential accuracy of magnetic localization.