Walking parameters estimation through channel state information preliminary results

TLDR: This paper has proposed a novel ubiquitous node deployment on the human body itself to minimize the environmental noise interference and trained the system to identify these signatures and deduce corresponding gait parameters.

Abstract: Stride rate and length are monitored by physicians to identify abnormalities in the patients' gait. These are also important parameters for athletes. Lab based video recordings are used for monitoring while new methods utilize inertial sensors. These sensors require a lot of sensed data processing consequently power is utilized. A new class of methods using wireless sensing has been introduced recently. These methods either require a lot of sensing nodes or are non-ubiquitous while others have used received signal strength as monitoring parameter which is quite unstable. We have proposed a novel ubiquitous node deployment on the human body itself to minimize the environmental noise interference. Walking parameters like stride rate, stride length have been estimated using wireless sensing physical layer channel state information (CSI). The human body acts as an obstacle for the wireless signals due to frequency selective multipath fading. We argue that this fading has a unique signature respective to the activity performed by humans which can be estimated using CSI. This signature should be best observed when both the sender and receiver nodes are deployed on the body due to decreased environmental interference. We have trained the system to identify these signatures and deduce corresponding gait parameters. In this paper we are only able to summarize the new idea with initial findings. We are in the process to understand the signal morphology for finer measurements.