Fundamental Investigation of Wi-Fi Beamforming Report Properties on Wireless Sensing

In the Wi-Fi network standards, channel state information (CSI) in IEEE 802.11n and compressed beamforming report (CBR) in IEEE 802.11ac and ax are utilized for high-speed and reliable wireless communications. Given that the CSI can accurately mirror physical changes in the wireless channel, CSI is also widely used for several sensing purposes. In recent years, CBR, the compressed form of CSI, is also used to realize wireless sensing by using IEEE 802.11ac-compatible wireless devices. Many studies on CSI-based sensing and CBR-based sensing rely on experimental measurements to demonstrate their feasibility, but the performance gap between the CSI-based and CBR-based sensing is not well discussed. In addition, the effect of the time-varying phase offset on the sensing performance is also still unclear. In this paper, we set up a simulation environment and clarify the feasibility of CBR-based sensing by dealing with practical sensing tasks with CBR. Our simulations find that CBR-based sensing with frequency analysis achieves high accuracy comparable to CSI-based sensing, whereas the time series data is considerably distorted due to bit quantization, and that existing sensing methods relying on the nature of CSI are unsuitable for CBR.