|
Compared with the traditional electronic reconnaissance receiver, the electronic reconnaissance receiving system using optical frequency comb down-conversion technology can simultaneously down-convert all the electromagnetic signals from the wide-band range (1MHz-20GHz) to the first repetition frequency interval range(DC-110MHz) for reception and measurement. This method can significantly improve the measuring speed and reduce the receiving cost. However, when the captured signal is a type with bandwidth like radar signal, the down-conversion spectrum may be aliased, resulting in the change of signal spectrum shape, which will affect the modulation type identification and frequency recovery of the signal. In order to solve the problem of modulation type identification of radar signals with spectrum aliasing after optical down-conversion, a modulation type identification method based on radar signal auto-correlation feature extraction is proposed in this paper. According to the characteristic that the auto-correlation function is insensitive to the spectrum aliasing after optical down-conversion, this method extracts the type identification features of auto-correlation envelope and its spectrum, such as skewness, kurtosis, impact factor, and finally uses SVM to identification the modulation type. Simulation results show that this identification method can accurately identify the modulation type of radar signal with spectrum aliasing after optical down-conversion, and has good anti-noise performance. When the SNR is higher than -3dB, the identification accuracy of this method for LFM and SIN after optical down-conversion is higher than 99%. For NP, BPSK, QPSK, BFSK, LFM and SIN signals with dynamic SNR between -6db and 10dB , the identification accuracy is more than 95%.
|
