Monitoring Human Head and Neck-Based Motions From Around-Neck Creeping Wave Propagations

The ability to monitor various activities related to the human head and neck is of critical importance to medical professionals. Radio-based wireless body-area networks provide a potential method of tracking human motion patterns without compromising the comfort or privacy of users. This letter investigates the feasibility of using the electromagnetic creeping wave propagation around the human neck to monitor neck-based motions. The channel around the human neck contains information regarding the neck and head movements; therefore, it can be considered for activity classification. The changes in the transmission coefficient (
S
₂₁
) between two on-body antennas are measured to classify four different head- and mouth-related motions including chewing, drinking, deep breathing, and speaking. To capture the time-varying signatures of the measured
S
₂₁
, the joint time–frequency analysis has been applied. The spectrograms are classified by a deep convolutional neural network.