Research On Identification Of The Radar Signal Based On The Phase Unwrapping

The identification of the radar signals based on the phase unwrapping in low Signal-Noise Ratio (SNR) is discussed in this dissertation.First, the phase and infrequency of the common radar signals are studied in this dissertation. The phase unwrapping algorithm can estimate the phase of the radar signals accurately when the SNR is no less than 6dB. But it is necessary to estimate the phase when the SNR is less than 6dB in fact. Thus the method of filtering the signals in the process of Hilbert transformation if the instantaneous bandwidth is narrow is put forward. It can be used to estimate the phase under the condition of low SNR. Consequently, the new phase unwrapping algorithm is applied to the frequency estimation of a sinusoid.Second, the identification of the radar signal modulation mode is also discussed in this dissertation, and a complete identification process is put forward. The time-phase and time-frequency curves of the signals are obtained by the improved phase unwrapping algorithm, and then an effective identification of the signals can be achieved by the characteristics of the time-phase and time-frequency curves. This process includes curve fitting, jump positions detection, peak detection, etc. The simulation results show that the recognition rate is greater than 90% when the SNR is not less than 3dB.In the end, the identification of S-Mode nonlinear chirp signals is discussed. First, the two characteristics of time-frequency are summarized, and then it is easy to identify the S-Mode nonlinear chirp signals based on these characteristics. In order to identify this signal in low SNR, a method of filtering the time-frequency curves before identifying the signals is put forward. This method can get a high rate of identification when then SNR is no less than 3dB.Lots of computer simulations prove that the method of filtering in the process of Hilbert transformation is benefit for the identification of the radar signals, and the performance of identification algorithm is good. Also the algorithm of identifying S-Mode nonlinear chirp signals is feasible and effective.