A coning micro-Doppler signals separation algorithm based on time-frequency information

Based on equivalent scatter center model of coning target, the micro-Doppler (m-D) signals induced by precession mainly consist of two components, i.e., cone node scatter signal varies in a sine and bottom scatter signal varies in a complex form time-frequency (TF) plane. As these two components are overlapped in the TF plane, this makes it hard to estimate the parameters of the target. Aiming at this problem, a novel m-D signal separation algorithm is proposed based on time-frequency information. As its sine variation and large frequency deviation benefit the precision of parameter estimation, the m-D signal of cone node scatter is separated by solving a Sinusoidal FM (SFM) signal parameter equation at first. Subsequently, the m-D signal with complex expression can be extracted from the TF plane by utilizing constant false alarm rate (CFAR) algorithm and centroid algorithm. Numerical results show the efficiency of the proposed algorithm.