Finger Gesture Sensing and Recognition Using a Wi-Fi-based Passive Radar

This paper presents a real-time finger gesture sensing system based on a passive radar that uses ambient Wi-Fi signals or those of an access point (AP). The core hardware of the system is an injection-locked quadrature receiver (ILQR) that receives the reference and echo signals of Wi-Fi for processing to obtain Doppler phase shifts for moving fingers. Additionally, a fast circumcenter estimation method is used for dc-offset compensation in the arctangent demodulation process. With the help of machine learning, gesture recognition can be carried out to distinguish accurately various finger gestures.