Abstract: With the growing demand for smart sensing applications, non-contact body detection technologies have received widespread attention due to their significant advantages in privacy protection and user experience. The analysis method based on commercial Wireless Fidelity (Wi-Fi) Channel State Information (CSI) provides a low-cost and sensorless scenario for smart home security monitoring and other scenarios by capturing the subtle perturbations caused by human activities on the multipath propagation of wireless signals. A classification model based on threshold logical judgment was constructed. The features of human body states in space were extracted through conventional algorithms, and the detection of human body states in space was achieved by using signal processing and logical judgment. Through real experiments, it was verified that an accuracy rate of over 99% was achieved in three different experimental scenarios respectively. This model can achieve good perception and detection in different scenarios, realizing high-precision perception of the human existence state in the environment.

Key words: Channel State Information, Contactless human detection, Wireless sensing, Smart home