Wi-CAS: A Contactless Method for Continuous Indoor Human Activity Sensing Using Wi-Fi Devices.

Area of Science:

• Computer Science
• Electrical Engineering
• Signal Processing

Background:

• Home isolation is crucial for virus containment but presents monitoring challenges.
• Traditional sensing methods are inadequate for large-scale, private indoor monitoring.
• Ubiquitous Wi-Fi offers a privacy-preserving, low-cost sensing alternative.

Purpose of the Study:

• To propose Wi-CAS, a contactless method for continuous indoor human activity sensing using Wi-Fi signals.
• To enable effective monitoring of individuals in home isolation.
• To develop a robust and low-cost solution for supervising isolated persons' activities.

Main Methods:

• Developed Wi-CAS, a contactless indoor person continuous activity sensing method.
• Utilized Wi-Fi signals for non-intrusive human activity detection.
• Implemented an ensemble classification method based on Hierarchical Clustering (HEC) for action recognition.

Main Results:

• Wi-CAS achieved a high combined activity recognition rate of 94.3%.
• The system demonstrated high sensitivity to unusual activities of individuals.
• Experimental evaluations in real home environments validated the method's effectiveness.

Conclusions:

• Wi-CAS offers a low-cost, highly robust solution for supervising home-isolated individuals.
• The proposed method effectively senses continuous movements and recognizes various actions.
• Wi-Fi-based sensing presents a viable approach for privacy-preserving indoor monitoring.