Abstract
Indoor magnetic-based positioning has attracted tremendous interests in recent years due to its pervasiveness and independence from extra infrastructure. Existing methods for indoor magnetic-based positioning are either point-based fingerprint matching or sequence-based fingerprint matching using the raw magnetic field strength. However, the magnetometers in smartphones are vulnerable to a few factors such as user's postures and walking speed, which causes the magnetic field strength corresponding to a location often shift in time or exhibit local distortions, thus greatly limits the positioning performance of existing methods rely on raw magnetic field strength. To this end, we observe the differences among magnetic field strength sequences are mainly attributed to small local segments, and design a new sequence-based fingerprint based on the differences among small local segments of raw MFS sequence to represent raw MFS sequence for indoor positioning. To demonstrate the utility of our proposed sequence-based fingerprint, we have performed a comprehensive experimental evaluation on two datasets, the results show that the proposed approach can significantly improve positioning performance compare with baseline methods.
Highlights
Recent years have witnessed an increasing attention on indoor positioning in view of its importance to indoor locationbased services, such as indoor advertising [1], patient activity monitoring [2] and indoor location recommendation [3]–[5]
We focus on indoor magnetic-based positioning and propose a novel sequence-based fingerprint based on the differences among small local segments of raw magnetic field strength (MFS) sequence, which can efficiently handle the local distortions and shift of raw MFS sequence
Each MFS sequence of training and test dataset is divided in several subsamples with 5 seconds for each one
Summary
Recent years have witnessed an increasing attention on indoor positioning in view of its importance to indoor locationbased services, such as indoor advertising [1], patient activity monitoring [2] and indoor location recommendation [3]–[5]. A few magnetic-based positioning methods [11]–[21] have been proposed in the literatures due to these special properties (for a review sees Section 2). Existing point-based fingerprint matching using the 3-D MFS vector makes no sense due to the following two factors: 1) the MFS at a given indoor location is a 3-D vector in space that varying with near location; 2) different orientation or postures of mobile phone lead to different MFS readings at the same location. Existing sequence-based fingerprint matching is far from satisfactory since the MFS values measured by mobile devices are vulnerable to external magnetic perturbations (e.g., the MFS may be distorted by metal or noise sources deriving from the surrounding environment), which often result in a magnetic distortion known as soft and hard iron effects
Sign-in/Register to view highlights & summary 
Published Version (
Free)
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
