Magnetic-Map-Matching-Aided Pedestrian Navigation Using Outlier Mitigation Based on Multiple Sensors and Roughness...

Area of Science:

• Robotics
• Geomatics Engineering
• Navigation Systems

Background:

• Pedestrian dead reckoning (PDR) using zero velocity updates (ZUPT) suffers from uncompensated position errors, leading to divergence.
• Accurate indoor localization is challenging due to the lack of easily obtainable position measurements, unlike in outdoor navigation.
• Existing indoor navigation methods require significant improvements in position accuracy.

Purpose of the Study:

• To propose a novel algorithm for accurate indoor pedestrian navigation.
• To enhance position accuracy in pedestrian dead reckoning by mitigating error divergence.
• To develop a magnetic field map-matching technique for reliable indoor localization.

Main Methods:

• Implemented a magnetic field map-matching technique using multiple magnetic sensors and importance sampling for position determination.
• Employed a normalization and roughness weighting method for outlier mitigation, improving sensor data reliability.
• Utilized a 15th-order error model and an importance-sampling extended Kalman filter for precise error correction in map-matching-aided pedestrian dead reckoning (MAPDR).

Main Results:

• The proposed magnetic field MAPDR algorithm demonstrated significant performance improvements in position accuracy compared to conventional PDR.
• Experimental results confirmed the effectiveness of the algorithm across various indoor environments.
• The integration of multiple sensors with advanced outlier mitigation enhanced the robustness of the navigation system.

Conclusions:

• The developed magnetic field MAPDR algorithm offers a substantial advancement for accurate indoor pedestrian navigation.
• The proposed method effectively compensates for position errors inherent in traditional PDR techniques.
• This research provides a viable solution for precise indoor localization using magnetic field mapping and advanced filtering.