Kalman/Map filtering-aided fast normalized cross correlation-based Wi-Fi fingerprinting location sensing.

Area of Science:

• Computer Engineering
• Signal Processing
• Wireless Communication

Background:

• Accurate indoor positioning remains a challenge.
• Existing Wi-Fi fingerprinting methods often rely on averaged signal data, limiting accuracy.
• The integration of advanced filtering techniques can improve localization performance.

Purpose of the Study:

• To propose a novel Wi-Fi fingerprinting location sensing system combining fast normalized cross correlation (FNCC) and Kalman/map filtering (KMF).
• To enhance fingerprinting accuracy by utilizing all received signal strength (RSS) samples and reference point variations.
• To improve the efficiency and accuracy of indoor positioning systems.

Main Methods:

• Developed a fast normalized cross correlation (FNCC) algorithm using all online RSS samples and reference point RSS variations.
• Implemented a Kalman/map filtering (KMF) approach incorporating a map matching algorithm to correct Kalman filter predictions based on indoor map structures.
• Utilized spatial proximities of consecutive localization results within the KMF.

Main Results:

• The FNCC algorithm demonstrated higher fingerprinting accuracy compared to conventional neighbor selection algorithms using RSS mean samples.
• FNCC achieved comparable accuracy to basic normalized cross correlation with reduced computational complexity.
• The KMF effectively improved location sensing accuracy by integrating indoor map information and spatial proximities, correcting prediction errors.

Conclusions:

• The proposed FNCC algorithm offers an efficient and accurate method for Wi-Fi fingerprinting.
• The KMF significantly enhances location sensing performance by leveraging indoor map constraints and sequential data.
• The combined FNCC-KMF system provides a robust solution for precise indoor positioning.