A Novel Valued Tolerance Rough Set and Decision Rules Method for Indoor Positioning Using WiFi Fingerprinting

Area of Science:

• Computer Science
• Electrical Engineering
• Geomatics

Background:

• WiFi fingerprinting is a key indoor positioning technique, but susceptible to environmental changes.
• Existing methods struggle with accuracy due to dynamic indoor conditions.
• Need for robust WiFi fingerprinting methods is critical for reliable indoor navigation.

Purpose of the Study:

• To propose a novel WiFi fingerprinting method robust to environmental changes.
• To enhance the accuracy and reliability of indoor positioning systems.
• To leverage rough set theory for improved WiFi-based localization.

Main Methods:

• Developed a valued tolerance rough set theory-based classification method for WiFi fingerprinting.
• Converted conventional Received Signal Strength (RSS) fingerprinting databases into decision tables.
• Constructed a new database with decision rules, including credibility degrees and support object set values.

Main Results:

• The proposed method achieved 98.05% accuracy in position classification.
• Demonstrated a significant improvement of approximately 50.49% over nearest-neighbor and random statistical methods.
• Reported mean positioning errors of 1.71 m and 1.99 m at wrongly estimated positions.

Conclusions:

• The novel WiFi fingerprinting method based on valued tolerance rough set theory offers superior performance.
• This approach provides a highly accurate and reliable solution for indoor positioning challenges.
• The method effectively addresses the vulnerability of traditional WiFi fingerprinting to environmental fluctuations.