Space-separated handwritten input technology based on commercial Wi-Fi devices
In this study, we designed a hand tracking scheme that only relied on signal processing method without modeling, which fundamentally solved the problem of poor generalization ability of the model.
MUSIC algorithm was used to estimate the initial position of AoA of dynamic signal arrival Angle, and the length change of dynamic reflection path was deduced according to the phase change of CSI. The Fresnel region intersection composed of two groups of transceiver equipment was used to calculate the hand coordinates. The median tracking error of 1.5cm was achieved in the perception area of 22.5m^2. Based on the above tracking technology, we have implemented a prototype system of blank handwriting input to provide reliable text input.
Passive 3D gesture tracking technology based on commercial Wi-Fi devices
Gestures actually contain 3D spatial information, and it is not accurate to only consider 2D plane estimation in gesture tracking and indoor track tracking systems, so we propose 3D gesture tracking technology.
In this study, the superresolution AOA-Tof MUSIC algorithm was used to extract the dynamic reflection path, the static signal component removal algorithm was proposed for tracking and calibration, and the gesture tracking model was designed in 3D space. The experimental results showed that the 3D median tracking error of our model in the 22.5m^2 perception area was up to 2.5cm.
A tracking scheme that does not depend on the initial position of the gesture(Researching)
In order to further apply the gesture tracking scheme to the practical application, a tracking scheme which does not require prior knowledge of the location of the receiving and receiving devices and does not depend on the initial position of the gesture is proposed. It uses the stability of arrival Angle and departure Angle in short time and the same phase bias between multipath signals to track the equipment, and finally achieves the millimeter tracking error in 2D space
 Zijun Han, Zhaoming Lu, Xiangming Wen, Jingbo Zhao, Lingchao Guo, Yue Liu. (2020). In-air handwriting by passive gesture tracking using commodity WiFi. IEEE Communications Letters, 24(11), 2652-2656.
 Zijun Han, Zhaoming Lu, Xiangming Wen, Wei Zheng, Jingbo Zhao, Lingchao Guo. CentiTrack: Towards Centimeter-Level Passive Gesture Tracking with Commodity WiFi, in IEEE Internet of Things Journal, (under review).
 Zijun Han, Lingchao Guo, Zhaoming Lu, Xiangming Wen, Wei Zheng. (2020, May). Deep adaptation networks based gesture recognition using commodity WiFi. In 2020 IEEE Wireless Communications and Networking Conference (WCNC) (pp. 1-7). IEEE.
 Zijun Han, Zhaoming Lu, Xiangming Wen, Lingchao Guo, Jingbo Zhao. Towards Sub-wavelength Level Passive 3D Gesture Tracking with Two WiFi Links. in IEEE Transactions on Mobile Computing, (under review) .