Research on Robust Adaptive RTK Positioning of Low-Cost Smart Terminals

Area of Science:

• Geomatics Engineering
• Navigation Systems
• Signal Processing

Background:

• Low-cost smart terminal Global Navigation Satellite System (GNSS) performance is constrained by hardware limitations and complex urban environments, impacting positioning accuracy and stability.
• Existing GNSS solutions struggle with signal interference and multipath effects common in built-up areas, necessitating advanced error mitigation techniques.

Purpose of the Study:

• To develop and evaluate a robust adaptive Kalman filter designed to enhance the positioning accuracy and stability of low-cost GNSS receivers, particularly in challenging urban settings.
• To mitigate the impact of outliers and system model errors on real-time kinematic (RTK) positioning performance for smartphones and dedicated low-cost receivers.

Main Methods:

• Implementation of a robust estimation approach within the Kalman filter algorithm to identify and down-weight abnormal GNSS observations.
• Integration of the Institute of Geodesy and Geophysics III (IGG III) function for adaptive regulation, modifying prior information using equivalent weight matrices and adaptive factors.
• Definition of a robust factor to dynamically adjust the weighting of positioning deviations based on pre- and post-robust estimates.

Main Results:

• Static experiments demonstrated significant improvements in positioning accuracy across multiple devices: Xiaomi 8 (29.6%), Huawei P40 (31.3%), Huawei mate40 (32.1%), and a low-cost M8 receiver (30.7%).
• Dynamic experiments yielded similar accuracy enhancements: Xiaomi 8 (28.3%), Huawei P40 (32.9%), Huawei mate40 (35.4%), and the M8 receiver (26.2%).
• A positive correlation was observed between smartphone positioning performance and robust RTK positioning; however, robustness effectiveness decreased at very high accuracy levels, suggesting increased sensitivity to interference.

Conclusions:

• The proposed robust adaptive Kalman filter effectively improves GNSS positioning accuracy and stability for low-cost smart terminals in various environments, especially urban areas.
• The method successfully reduces the influence of outliers and system model errors, enhancing the reliability of RTK positioning.
• While beneficial, the diminishing robustness at high accuracy levels indicates a need for further research into specialized interference mitigation for high-precision applications.