A Low-Complexity CSI-Based Wifi Sensing Scheme for LoS-Dominant Scenarios

Integrating sensing into wifi platforms, referred to as wifi sensing, provides an efficient and device-free means for indoor monitoring/localization with low cost. Clock asynchrony is one of the most challenging issues in wifi sensing. A mainstream solution to date employs cross-antenna processing to suppress the clock offsets that are common to all antennas. Such methods, however, may suffer from issues such as mirrored targets and noise enhancement etc. This paper develops a novel wifi sensing scheme. It embodies accurate and low-complexity methods for estimating the timing and frequency offsets as well as the angle-of-arrival (AoA), all from the LoS path. It also involves a coherent Doppler processing method which effectively suppresses static paths and accurately estimates the Doppler by enjoying the coherent processing gain. Corroborated by experimental results using open Widar2.0 data, the proposed design is able to precisely recover the velocity traces in various indoor scenarios.