CSHA-CSI: Towards Contactless Sensing of Human Activities Utilizing WiFi Channel State Information

Abstract

Numerous studies have demonstrated the interest of reusing wireless signals (such as WiFi) for detecting human activity and therefore realizing numerous innovative applications, including security surveillance, daily activity tracking, monitoring vital signs, and identifying users, which require more precise motion detection. Such solutions may facilitate a broad spectrum of uses for smart homes and offices, encompassing security, healthcare monitoring and management, intelligent healthcare solutions, and smart appliance interaction. Through the analysis of the wireless signals that have been received, it is possible to detect human motions due to the effects that human movements exert on the transmission of wireless signals (including phenomena like reflection, diffraction, and scattering). Wireless sensing of human activity has gained significant interest in recent years and emerged as a critical field of research due to its low cost and non-intrusiveness. In this study, we introduce a non-invasive as well as robust system to recognize human activity, utilizing affordable and widely used WiFi devices, eliminating the need for individuals to don or transport items. Our system also has the capability for extensive deployment and continuous longterm monitoring. Experimental results demonstrate the reliability as well as the precision of our system within realistic settings. This is a clear indication that non-invasive, continuous fine-grained human activity monitoring is feasible without incurring additional costs.