Ding Yi

Ding Yi

Associate Professor in GNSS and Multisensor Navigation

York-Huawei Smartphone Collaboration Project

Project Overview

This project focused on the development and evaluation of advanced smartphone-based GNSS positioning algorithms for realistic vehicle navigation scenarios. The primary objective was to design and implement core RTK and PPP GNSS+IMU C++ processing software, Jaqen, enabling high-accuracy positioning using low-cost smartphone sensors in challenging driving environments.

An end-to-end framework covering algorithm development, field experimentation, and performance validation was established through close collaboration between York University and Huawei.

Phase 1: RTK GNSS + IMU Algorithm Development

  • Developed multi-GNSS (GPS, GLONASS, Galileo, BeiDou) single- and dual-frequency float RTK software using a Sequential Least Squares (SLS) estimator.
  • Designed an Extended Kalman Filter (EKF) architecture supporting both loosely coupled and tightly coupled RTK+IMU integration.
  • Implemented robust measurement modeling and state propagation tailored to smartphone-grade GNSS and IMU sensors.

Phase 2: Vehicle Field Testing and RTK+IMU Evaluation

  • Customized pre-processing and post-processing modules to accommodate smartphone-specific GNSS characteristics.
  • Organized and conducted vehicle field tests across diverse environments, including:
    • Open-sky highways
    • Vegetation-covered roads
    • Moderate urban environments
    • Overpasses and complex road geometries
  • Evaluated positioning performance in simulated real-time mode, achieving acceptance criteria for:
    • 95th and 68th percentile horizontal accuracy
    • Positioning success rate

Phase 3: PPP GNSS + IMU Algorithm Development

  • Extended the RTK+IMU framework by integrating Precise Point Positioning (PPP) functionalities.
  • Developed multi-GNSS single- and dual-frequency float PPP processing modules with an ionospheric-constrained strategy.
  • Participated in the 2022 Google Smartphone Decimeter Challenge (GSDC), achieving a score of 2.485 using simulated real-time solutions.

Phase 4: Vehicle Field Testing and PPP+IMU Evaluation

  • Enhanced traditional PPP solutions by incorporating code-only satellites and measurements.
  • Led large-scale vehicle measurement campaigns and validated PPP+IMU performance under realistic driving conditions.
  • Successfully met the same acceptance criteria as RTK+IMU testing for:
    • Horizontal accuracy percentiles
    • Positioning robustness and reliability
  • Coordinated York–Huawei technical meetings and led the submission of final project deliverables.

Key Outcomes

  • End-to-end C++ GNSS+IMU processing software for smartphone-based RTK and PPP applications.
  • Demonstrated Accurate positioning performance (<1.5 m>) in realistic driving environments using low-cost sensors.
  • Provided a solid technical foundation for subsequent research on smartphone PPP, multi-sensor integration, and urban navigation.