SDR-based concurrent multi-band event matching using bistatic Wi-Fi "radar"

As a significant component of identification, the requirements for Human Activity Recognition (HAR) are rising in the fields of healthcare, entertainment and leisure. Recent years, Wi-Fi, as a ubiquitous Radio Frequency (RF) signal, is widely used in sensing fields, including HAR. In this thesis, we proposed a concurrent dual-band event matching system, which is based on Software-Defined Radio (SDR). Compared with traditional wearable sensors, privacy issues are solved. And unlike Wi-Fi commercial equipment, finer information can be extracted. We identify a set of events through Channel State Information (CSI) and Doppler profile extracted from Wi-Fi signals. Experimental results show that dual-band can improve event matching accuracy and can achieve an average event-matching accuracy of 98.54%.