Feature extraction of orbital shift frequency signal based on time-frequency analysis

Traditional signal analysing is difficult to meet the detection demands of Orbital shift frequency signals effectively. This paper proposes a new scheme on the basis of time-frequency analysis to detect the frequency-shift signals. After analyzing the acquirement and spectrum of the orbital signals, the spectrum refining algorithm based on under-rate sampling and ZFFT is put forward to improve the accuracy of low frequency. And then, the wavelet ridge line technique is used to calculate the upper and down side frequencies to achieve the signals´ feature extraction.