A time-frequency analysis method with joint speed estimation and translation compensation based on Near-Field MIMO Array

Micro Doppler analysis of human kinematics is widely used in the detection and recognition of human contraband. However, during the observation time (CPI), the translational motion of the human body introduces errors in the Doppler curve of the micro moving parts of body, thereby altering the Doppler dynamic range and causing folding and deviation of the micro Doppler curve. At this point, conducting time-frequency analysis directly on the echo data will result in significant errors. This paper proposes a joint algorithm of translational speed estimation and time-frequency analysis based on the MIMO near-field model. This algorithm not only achieves the estimation of translational speed, but also compensates for the phase error in the micro motion part of body, thereby improving the recognition performance of time-frequency maps. The effectiveness and superiority of this algorithm have been verified through simulation.