Multi-Stage Pedestrian Positioning Using Filtered WiFi Scanner Data in an Urban Road Environment

Area of Science:

• Wireless Sensor Networks
• Indoor and Outdoor Localization
• Signal Processing

Background:

• Wireless sensor networks (WSN) are increasingly used for localization.
• Received Signal Strength Indicator (RSSI) from WiFi is a common method for low-speed traffic positioning.
• Existing methods often struggle with accuracy in complex urban environments.

Purpose of the Study:

• To develop a novel range-based localization method for high-accuracy outdoor pedestrian positioning.
• To improve upon existing Received Signal Strength Indicator (RSSI) based localization techniques.
• To address the challenges of urban road environments for accurate positioning.

Main Methods:

• Offline stage: Piecewise Polynomial Regression Model (PPRM) for Euclidean distance estimation, replacing traditional propagation models (PM).
• Online stage: Constant Velocity Kalman Filter (CVKF) for RSSI smoothing and distance estimation, Least Squares Taylor Series Expansion (LS-TSE) for 2D coordinate calculation, and Unscented Kalman Filter (UKF) for trajectory smoothing.
• Utilized field data from Guangzhou, China for experimental validation.

Main Results:

• The combined CVKF and PPRM achieved a highly accurate distance estimator with an error of <1.98 m at 90% probability, outperforming existing propagation models.
• The online method demonstrated an average positioning error of 1.67 m.
• The proposed method significantly outperforms classical localization techniques.

Conclusions:

• The developed method offers a significant advancement in outdoor pedestrian positioning accuracy using WiFi RSSI.
• The combination of PPRM and CVKF provides a robust and accurate distance estimation.
• The system is effective in real-world urban road environments, offering a superior alternative to existing methods.