Analysis of Radar Doppler Signature from Human Data

Abstract

This paper presents the results of time (autocorrelation) and time-frequency (spectrogram) analyses of radar signals returned from the moving human targets. When a radar signal falls on the human target which is moving toward or away from the radar, the signals reflected from different parts of his body produce a Doppler shift that is proportional to the velocity of those parts. Moving parts of the body causes the characteristic Doppler signature. The main contribution comes from the torso which causes the central Doppler frequency of target. The motion of arms and legs induces modulation on the returned radar signal and generates sidebands around the central Doppler frequency, referred to as micro-Doppler signatures. Through analyses on experimental data it was demonstrated that the human motion signature extraction is better using spectrogram. While the central Doppler frequency can be determined using the autocorrelation and the spectrogram, the extraction of the fundamental cadence frequency using the autocorrelation is unreliable when the target is in the clutter presence. It was shown that the fundamental cadence frequency increases with increasing dynamic movement of people and simultaneously the possibility of its extraction is proportional to the degree of synchronization movements of persons in the group.