A Non Intrusive Human Presence Detection Methodology Based on Channel State Information of Wi-Fi Networks

Area of Science:

• Electrical Engineering
• Computer Science
• Signal Processing

Background:

• Indoor localization services are increasingly important for security and personalized applications.
• Ubiquitous Wi-Fi network deployment enables novel sensing capabilities.
• Demand for non-intrusive presence detection systems is growing in various environments.

Purpose of the Study:

• To develop a non-intrusive human presence detection methodology.
• To leverage Channel State Information (CSI) from 802.11n wireless networks for presence detection.
• To validate the proposed methodology's effectiveness in real-world scenarios.

Main Methods:

• Utilized Channel State Information (CSI) from 802.11n wireless communication networks.
• Performed time-domain analysis and feature extraction on CSI data.
• Trained and validated classification models using captured data from diverse environments.

Main Results:

• The proposed methodology successfully detects human presence using Wi-Fi CSI.
• Classification models achieved an average accuracy exceeding 90%.
• The system demonstrated effectiveness across different indoor environments.

Conclusions:

• CSI-based presence detection offers a viable non-intrusive solution.
• The 802.11n standard, with MIMO and OFDM, provides rich data for this application.
• This approach has significant potential for enhancing indoor security and services.