A Self-Adaptive Model-Based Wi-Fi Indoor Localization Method.

Area of Science:

• Computer Science
• Electrical Engineering
• Ubiquitous Computing

Background:

• Existing indoor localization methods often require specialized hardware, frequent recalibration, or are not suitable for long-term deployment.
• These limitations hinder the practical application of many indoor positioning technologies in real-world scenarios.

Purpose of the Study:

• To present a novel, self-calibrating, and self-adaptive indoor localization method utilizing only Wi-Fi signals.
• To develop a maintenance-free localization system suitable for real-world deployments without requiring additional hardware beyond standard Wi-Fi access points.

Main Methods:

• The method employs two propagation models (free space path loss and ITU) extended with additional parameters for enhanced simulation.
• A self-calibrating procedure uses one model to infer spatial parameters and the other to simulate signal propagation.
• It operates without requiring the mobile terminal in access-point mode, relying solely on Wi-Fi access point and environmental (wall) data.

Main Results:

• Evaluated in single- and multi-room environments, the method achieved mean errors of 2-3 meters and 3-4 meters, respectively.
• Performance is comparable to existing indoor localization techniques.
• Demonstrated usable localization accuracy using a Wi-Fi-only approach with minimal hardware and software requirements.

Conclusions:

• The proposed Wi-Fi-based method offers a practical and accurate solution for indoor localization in real-world settings.
• Its self-calibrating and self-adaptive nature eliminates the need for maintenance and specialized hardware.
• This approach significantly lowers the barrier for deploying effective indoor positioning systems.