New signal location method based on signal-range data for proximity tracing tools.

Area of Science:

• Computer Science
• Signal Processing
• Geometry

Background:

• Accurate indoor signal localization is crucial for technologies like COVID-19 proximity contact tracing.
• Previous work introduced five geometric methods for signal localization, including the Pole-Polar Centroid (PPC) model.
• The PPC model, while precise, exhibited lower accuracy compared to other geometric methods.

Purpose of the Study:

• To propose an enhanced signal localization method, the weighted Pole-Polar Centroid (wPPC), to improve accuracy.
• To address the limitations of the existing PPC method in real-world scenarios with signal data errors.

Main Methods:

• Developed the weighted Pole-Polar Centroid (wPPC) method, an extension of the original PPC model.
• Utilized geometric principles for signal triangulation, maintaining O(m^2) complexity and minimum dimensionality of m=2 nodes.
• Tested the wPPC method using an IEEE 802.11 network infrastructure for signal interaction and location estimation.

Main Results:

• The wPPC method demonstrated improved accuracy in device location estimation compared to the original PPC method.
• wPPC effectively mitigated the influence of common signal data errors, leading to more reliable results.
• The proposed method outperformed other geometric approaches and conventional numeric methods in accuracy.

Conclusions:

• The weighted Pole-Polar Centroid (wPPC) method offers a significant advancement in range-based signal triangulation.
• This method is applicable to various signal types beyond IEEE 802.11, including Wi-Fi, Bluetooth, light, and sound.
• wPPC provides a robust solution for accurate indoor localization, essential for proximity tracing and other applications.