INSTANTANEOUS FREQUENCY ESTIMATION METHODS OF MICRO-DOPPLER SIGNAL

Extracting instantaneous frequency (IF) of Micro-Doppler (M-D) is the key to estimating micro-motion parameters. In this paper, firstly, the micro-motion model of coning was set up. Meanwhile, the theoretical analysis and derivation of micro-Doppler were performed. Then we introduced occlusion effect and Time Frequency transform to better accord with the complexity and reality of feature extraction of Micro-Doppler in the real world. Besides other computational complex separation methods, we extended the Viterbi algorithm for some cases and proposed two novel means, selecting local maximum value and extracting central position, to solve the dilemma where the Viterbi algorithm failed. Finally corresponding simulation corrected the performance of the two methods in different Signal to Noise Ratio (SNR). The simulation analysis showed that the methods were effective if time-frequency map was not polluted seriously by noise.