Automatic Detection of Changes in Signal Strength Characteristics in a Wi-Fi Network for an Indoor Localisation...

Area of Science:

• Computer Science
• Electrical Engineering
• Signal Processing

Background:

• Wi-Fi-based indoor localization systems are susceptible to reduced Quality of Service (QoS) due to changes in Received Signal Strength (RSS).
• Variations in RSS from Access Points (APs) can stem from network reconfigurations, malfunctions, or infrastructure aging.

Purpose of the Study:

• To propose a dynamic system that estimates and compensates for changes in AP RSS to maintain localization accuracy.
• To enhance the robustness and reliability of Wi-Fi-based indoor localization systems.

Main Methods:

• Developed a dynamic system employing an RSS estimator that utilizes readings from multiple APs.
• Implemented an optimal threshold algorithm to detect APs exhibiting characteristic changes.
• Rebuilt the localization model by excluding identified aberrant APs to preserve QoS.

Main Results:

• Simulated RSS changes by relocating AP models within a multi-floor building environment.
• Evaluated system performance in isolated single-floor, single-floor, and multi-floor building simulations.
• Demonstrated a mean increase in localization error of 0.25–0.61 m with the system, compared to 1.21–1.98 m without it.

Conclusions:

• The proposed dynamic system effectively mitigates the impact of RSS variations on Wi-Fi localization accuracy.
• The system offers a robust solution applicable to various Wi-Fi services facing infrastructure changes, malfunctions, or aging.