Research Projects

We do not need sensors, wearables but only the exisiting WiFi deployment together with our AI Techniques, Signal Process Methods, Radar Techniques, 5G Communication Skills, Massive MIMO Knowledge, mmWave Studies… to make your life easier and convenient, by your control of course . Know more about Wireless AI for wireless sensing and localization using ambient radio signals…

AI Techniques for Wireless Sensing

• We created the first-of-its-kind device-free motion sensing system which was commercialized by Origin Wireless AI Inc. and won the CES 2020 Innovation Award. In partnership with Belkin, it has been deployed over 150 countries worldwide as LinkSysAware.
  

• We developed an accurate Sleep Monitoring System and commercialized it as Origin Health for Remote Patient Monitoring by Origin Wireless AI Inc., which won the CES 2021 Best of Innovation Award.
  

• We designed a Driver’s Arrival Time Sensing System, which makes the car more smart in multiple-service management than the existing driver’s presence sensing only system. Real-world tests show 92.5% accuracy with less than 1s error. (PDF)
  

• We developed a Millimeter Wave (mmWave) based Vital Sign Monitoring System which can accurately estimate the breathing rate, heart rate and heart rate viability (HRV) for multiple objects. Part of the system has been invited to demo for Qualcomm and it is being commercialized… (PDF）
  

• We developed a massive MIMO based localization and tracking system on the base of Time Reversal techniques, which can benifit from the unavoidable multipath distortions in our daily life and thus improves the scalability greatly. (PDF）
  

• We developed a singal decomposition method which can extract the intrinsic signal components over multiple channels simultaneously while does not need the prior information about the number of the signal components. It can be widely applied becasue multiple channel data such as EEG data are routinely acquired owing to the development of sensor array techniques.
  

• We developed an Time Reversal based localization methods which can guide the transmitting signal to focus on the target and thus improves the SNR up to 5dB and the localization accuracy. (PDF）
  

• We establish a link between the autocorrelation function (ACF) of the physical layer CSI and the speed of a moving object. We then developed an active and a passive Indoor Tracking Systems with decimeter-accuracy. The research facilitated three ongoing products including Cart-tracking in supermarkets, Fall-detector and Gait-recognition, which has been partly invited for demo to Apple, HP, Qualcomm, Verizon, Marvell.
  

• DOA Estimations using different spectral analysis methods
  

• Automatic Test System Based on LabView