On the Dynamic RSS Feedbacks of Indoor Fingerprinting Databases for Localization Reliability Improvement

Area of Science:

• Ubiquitous Computing
• Context-Aware Applications
• Indoor Positioning Systems

Background:

• Received Signal Strength (RSS) is vital for indoor localization.
• Existing RSS fingerprinting databases are costly and impractical for dynamic environments.
• High-traffic areas present unique challenges for maintaining accurate location data.

Purpose of the Study:

• To develop a dynamic estimation resampling method for indoor positioning.
• To reduce the cost and improve the efficiency of RSS fingerprinting databases.
• To enhance the accuracy of location-aware applications in changing environments.

Main Methods:

• Proposing a dynamic estimation resampling method.
• Adaptively applying updated and offline fingerprinting points based on location.
• Utilizing temporal and spatial strength for resampling.

Main Results:

• Achieved double correctness probability with only 3% feedback in simulations.
• Demonstrated excellent 1-meter accuracy errors in real-world environments.
• The method is efficient and cost-effective for dynamic settings.

Conclusions:

• The dynamic estimation resampling method offers a practical solution for indoor positioning.
• It overcomes the limitations of traditional RSS fingerprinting in changing environments.
• This approach enhances the reliability and accuracy of location-based services.