S-PDR: A Novel Pedestrian Dead Reckoning Algorithm with step-based attitude corrections for Free-Moving Handheld devices
Date
2021-01-08
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
Mobile location-based services (MLBS) are attracting attention for their potential public applications and personal use. MLBS can be used for a variety of applications such as location-based advertisement, smart shopping, smart cities, health applications, emergency response, and even gamming. The majority of these applications are used in indoor environments where the well established GNSS navigation solutions are hindered or even unavailable and hence they rely on alternative navigation solutions such Inertial Navigation Systems (INS). To date, the main challenges for MLBS is to provide accurate and reliable navigation solution under varying circumstances such as indoor or outdoor, while reducing system cost and having real-time applicability, which is achieved through the use of MEMS technology. However, MEMS sensors suffer from high errors and noise to signal ratio that results in quick divergence of the INS solution, hence the need for aiding. This thesis aims at providing a Pedestrian Dead Reckoning (PDR) solution that uses off-the-shelf sensors in mobile devices to provide short term reliable navigation solution that helps reduce the complexity and frequency of relying on aiding techniques through developing a novel PDR system S-PDR . S-PDR utilizes a novel step detection technique that is motion-mode and use-case invariant, an attitude correction technique that can provide corrections as frequently as a step-by-step basis, and an enhanced PCA-based heading estimation. Testing results in comparison to XSense MTi G-710 which is a high-end MEMS sensor show that S-PDR provide reliable short-term navigation solution with final positioning error that is up to 6 meters after 3 minutes operation time, outperforming the on-board fusion solution provided by the XSense. The short term enhancement of the PDR solution reliability can help reduce the operational complexity of aiding navigation systems such as RF-based indoor navigation and Magnetic Map Matching as it reduces the frequency by which these aiding techniques are required and applied.
Description
Keywords
Pedestrian Dead Reckoning, Attitude correction, Step detection, Heading Estimation, Least-Squares Estimation, Complementary Filter, Sensor fusion, Smartphone
Citation
Khedr, M. E. (2020). S-PDR: A Novel Pedestrian Dead Reckoning Algorithm with step-based attitude corrections for Free-Moving Handheld devices (Doctoral thesis, University of Calgary, Calgary, Canada). Retrieved from https://prism.ucalgary.ca.