Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Pedestrian Inertial Navigation: An Overview of Model and Data-Driven Approaches

Itzik Klein

TL;DR

This work surveys indoor walking navigation using inertial sensors, contrasting shoe-mounted INS and PDR with unconstrained devices. It presents a model-based PDR framework with step detection, step-length estimation, heading determination, and position update, and augments it with SM-INS concepts such as EKF-based fusion, zero-velocity detection, and information aiding. It then surveys data-driven PDR strategies—activity-assisted, hybrid, and learning-based frameworks—covering neural-network basics, step-length/heading regression, and end-to-end position estimation. The results emphasize how SM-INS with information aiding dramatically reduces drift, while data-driven approaches offer robust alternatives for activity context, heading, and step-length estimation, with significant implications for accurate, GNSS-free indoor navigation.

Abstract

The task of indoor positioning is fundamental to several applications, including navigation, healthcare, location-based services, and security. An emerging field is inertial navigation for pedestrians, which relies only on inertial sensors for positioning. In this paper, we present inertial pedestrian navigation models and learning approaches. Among these, are methods and algorithms for shoe-mounted inertial sensors and pedestrian dead reckoning (PDR) with unconstrained inertial sensors. We also address three categories of data-driven PDR strategies: activity-assisted, hybrid approaches, and learning-based frameworks.

Pedestrian Inertial Navigation: An Overview of Model and Data-Driven Approaches

TL;DR

This work surveys indoor walking navigation using inertial sensors, contrasting shoe-mounted INS and PDR with unconstrained devices. It presents a model-based PDR framework with step detection, step-length estimation, heading determination, and position update, and augments it with SM-INS concepts such as EKF-based fusion, zero-velocity detection, and information aiding. It then surveys data-driven PDR strategies—activity-assisted, hybrid, and learning-based frameworks—covering neural-network basics, step-length/heading regression, and end-to-end position estimation. The results emphasize how SM-INS with information aiding dramatically reduces drift, while data-driven approaches offer robust alternatives for activity context, heading, and step-length estimation, with significant implications for accurate, GNSS-free indoor navigation.

Abstract

The task of indoor positioning is fundamental to several applications, including navigation, healthcare, location-based services, and security. An emerging field is inertial navigation for pedestrians, which relies only on inertial sensors for positioning. In this paper, we present inertial pedestrian navigation models and learning approaches. Among these, are methods and algorithms for shoe-mounted inertial sensors and pedestrian dead reckoning (PDR) with unconstrained inertial sensors. We also address three categories of data-driven PDR strategies: activity-assisted, hybrid approaches, and learning-based frameworks.
Paper Structure (19 sections, 38 equations, 17 figures, 2 algorithms)

This paper contains 19 sections, 38 equations, 17 figures, 2 algorithms.

Table of Contents

  1. Introduction
  2. Model-Based Pedestrian Dead Reckoning
  3. Step detection approaches
  4. Step-length estimation
  5. Heading and walking direction
  6. Two-dimensional positioning
  7. Three-dimensional PDR
  8. Shoe-Mounted INS
  9. Inertial Navigation
  10. Navigation Filter
  11. Zero Velocity Detectors
  12. Information Aiding
  13. Summary and Analytical Assessment
  14. Data-Driven Pedestrian Dead Reckoning
  15. A Brief Introduction to Neural Networks
  16. ...and 4 more sections

Figures (17)

  • Figure 1: Shoe-mounted INS (left) requires solving the INS equations and a nonlinear filter with information aiding. The PDR framework (right) is applicable to unconstrained inertial sensors, such as those in smartphones.
  • Figure 2: Model-based PDR stages.
  • Figure 3: Specific force magnitude as a function of time and its average value. This recording was taken with a smartphone in texting mode by a pedestrian walking for 25.4 meters.
  • Figure 4: Identified user steps. This recording was taken with a smartphone in texting mode by a pedestrian walking for 25.4 meters.
  • Figure 5: Estimated user distance of the three step-length approaches. The recording was obtained with a smartphone in texting mode by a pedestrian walking for 25.4 meters.
  • ...and 12 more figures