Method for Identifying Intra-Pulse Modulation Type of Frequency Modulation Signal by Combining Short-Time Fourier Transform and Phase Difference Instantaneous Frequency

Abstract

The present application provides a method for identifying the in-pulse modulation type of a frequency modulation signal by combining the short-time Fourier transform and the phase difference instantaneous frequency. Step 1: Calculate the instantaneous frequency of the intermediate frequency sampled signal by using the short-time Fourier transform method; Step 2: Calculate the bandwidth of the signal; Step 3: If it is less than the set threshold, it is an unmodulated signal; otherwise, it is a frequency modulation signal and enters Step 8; Step 4: Calculate the instantaneous frequency by using the phase difference method; Step 5: Obtain the instantaneous frequency of the signal and calculate the signal bandwidth; Step 6: If the signal bandwidth is greater than the set threshold, enter Step 7; otherwise, it is an unmodulated signal; Step 7: If the error after linear fitting is greater than the set threshold, it is a non-linear frequency modulation signal; otherwise, it is a linear frequency modulation signal; Step 8: Fit the instantaneous frequency curve and calculate the error after fitting. If the error after fitting is greater than the set threshold, it is a non-linear frequency modulation signal; otherwise, it is a linear frequency modulation signal. The present application combines the advantages of the short-time Fourier transform and the phase difference instantaneous frequency calculation method.

Description

Technical Field

[0001] This application relates to the technical field of intra-pulse modulation types, and particularly to a method for identifying the intra-pulse modulation type of a frequency-modulated signal by combining short-time Fourier transform and phase-difference instantaneous frequency. Background Art

[0002] In current intra-pulse modulation identification methods, generally, a method of making a decision tree decision after feature extraction is used for identification. When identifying the intra-pulse modulation of a frequency-modulated signal, the time-bandwidth product can be used to distinguish the frequency-modulated signal from other types of signals. To further distinguish between linear frequency-modulated signals and non-linear frequency-modulated signals, it is necessary to calculate the instantaneous frequency modulation curve of the signal and distinguish it through the straight-line fitting error. Typical methods for calculating the instantaneous frequency curve are short-time Fourier transform and phase difference.

[0003] Both of the above two instantaneous frequency calculation methods have defects. The short-time Fourier transform method has low frequency resolution and cannot distinguish frequency-modulated signals with small bandwidths. The phase-difference calculation method will make mistakes in phase deblurring under large bandwidth conditions, resulting in incorrect calculation of the instantaneous frequency curve. This leads to incorrect identification of the intra-pulse modulation type of the frequency-modulated signal.

[0004] Therefore, using only one of the above instantaneous frequency calculation methods alone cannot adapt to the identification of the intra-pulse modulation type of frequency-modulated signals with all bandwidths. Summary of the Invention

[0005] This application provides a method for identifying the intra-pulse modulation type of a frequency-modulated signal by combining short-time Fourier transform and phase-difference instantaneous frequency, which can be used to solve the technical problem that the existing methods cannot adapt to the identification of the intra-pulse modulation type of frequency-modulated signals with all bandwidths.

[0006] This application provides a method for identifying the intra-pulse modulation type of a frequency-modulated signal by combining short-time Fourier transform and phase-difference instantaneous frequency. The method includes:

[0007] Step 1: Calculate the instantaneous frequency of the intermediate-frequency sampled signal S(n) using the short-time Fourier transform method;

[0008] Step 2: Calculate the bandwidth B of the signal through the instantaneous frequency curve;

[0009] Step 3: If τB is less than the set first threshold, it is an unmodulated signal or a frequency-modulated signal with a small bandwidth; otherwise, it is a frequency-modulated signal; go to Step 8;

[0010] Step 4: For an unmodulated signal or a frequency-modulated signal with a small bandwidth, calculate the instantaneous frequency using the phase-difference method;

[0011] Step 5: Obtain the bandwidth of the signal by obtaining the instantaneous frequency of the signal using the phase difference;

[0012] Step 6: If the signal bandwidth is greater than the set second threshold, go to Step 7; otherwise, it is judged as an unmodulated signal.

[0013] Step 7: If the error after fitting is greater than the set third threshold, it is judged as a non-linear frequency modulation signal; otherwise, it is judged as a linear frequency modulation signal.

[0014] Step 8: Perform a linear fit on the instantaneous frequency curve obtained by the short-time Fourier transform and calculate the error after fitting.

[0015] Go to Step 9;

[0016] Step 9: If the error after fitting is greater than the set fourth threshold, it is judged as a non-linear frequency modulation signal; otherwise, it is judged as a linear frequency modulation signal.

[0017] Furthermore, use the short-time Fourier transform method to calculate the instantaneous frequency of the intermediate-frequency sampled signal S(n), including:

[0018] Step 11: Divide the intermediate-frequency sampled signal into N / 2 overlapping signal segments and denote them as S m (n);

[0019] where N is the number of points of the short-time Fourier transform; m is the m-th segment, and n is the n-th sampling point;

[0020] Step 12: Calculate the fast Fourier transform of S m (n) segment by segment, and the expression is as follows:

[0021] F m (n) = fft(S m (n)), n = 0:N-1

[0022] Step 13: Calculate the frequency corresponding to the maximum power of the fast Fourier transform segment by segment:

[0023] f m (n) = maxp(fabs(F m (n))) * Fs / N, where Fs is the signal sampling frequency

[0024] Step 14: After the calculation, obtain the instantaneous frequency curve f(m).

[0025] Furthermore, in Step 2, the bandwidth B is determined by the following method:

[0026] B = max(f m (n)) - min(f m (n))

[0027] According to the pulse width τ of the signal, calculate the time-bandwidth product τB of the signal.

[0028] Further, for an unmodulated signal or a small-bandwidth frequency modulation signal, the instantaneous frequency is calculated using the phase difference method, including:

[0029] Step 41, perform an orthogonal transformation on the intermediate-frequency sampled signal S(n):

[0030] S'(n) = hilbert(S(n))

[0031] Step 42, calculate the phase of the signal using the orthogonal signal;

[0032] θ(n) = atan2(imag(S'(n)), real(S'(n)))

[0033] Step 43, obtain the instantaneous frequency of the signal using the phase difference:

[0034] f(m) = Fs(θ(n) - θ(n - 1)) / (2π), where Fs is the signal sampling frequency.

[0035] Further, in step 5, the signal bandwidth is determined by the following method:

[0036] B = max(f(m)) - min(f(m))

[0037] Perform a linear fit on the instantaneous frequency curve and calculate the error after fitting.

[0038] In this application, the method of using the short-time Fourier transform to calculate the instantaneous frequency of a signal has good anti-noise ability but limited frequency resolution and cannot effectively identify the in-pulse modulation signal with small-bandwidth modulation. On the basis of having been determined as small-bandwidth modulation, the phase difference method is further used to calculate the instantaneous frequency of the signal to improve the frequency resolution of the instantaneous frequency calculation; on this basis, a refined determination of unmodulated signals and frequency modulation signals is carried out. After testing, the method provided in this application can adapt to the in-pulse modulation signal type recognition of frequency modulation signals with a bandwidth of 1 MHz and Fs / 2 bandwidth. Description of the Drawings

[0039] Figure 1 It is a general block diagram of a method for identifying the in-pulse modulation type of a frequency modulation signal by combining the short-time Fourier transform and the instantaneous frequency of the phase difference provided by an embodiment of this application;

[0040] Figure 2 It is an instantaneous frequency curve diagram of the short-time Fourier transform of a sinusoidal frequency modulation signal provided by an embodiment of this application. Detailed Embodiments

[0041] To make the objectives, technical solutions, and advantages of this application clearer, the embodiments of this application will be further described in detail below in conjunction with the accompanying drawings.

[0042] By combining the advantages of the short-time Fourier transform and the phase-difference instantaneous frequency calculation method; then constructing a decision tree to distinguish unmodulated signals, linear frequency modulation signals, and non-linear frequency modulation signals. The overall block diagram of a method for identifying the in-pulse modulation type of frequency modulation signals combining the short-time Fourier transform and the phase-difference instantaneous frequency is as Figure 1 shown.

[0043] Step 1: Use the short-time Fourier transform method to calculate the instantaneous frequency of the intermediate-frequency sampled signal S(n). The calculation steps are as follows.

[0044] Step 11: Divide the intermediate-frequency sampled signal into N / 2 overlapping signal segments and denote them as S m (n);

[0045] where N is the number of points of the short-time Fourier transform. Here, m is the m-th segment and n is the n-th sampling point;

[0046] Step 12: Calculate the fast Fourier transform of S m (n) segment by segment. The expression is as follows:

[0047] F m (n) = fft(S m (n)), n = 0:N-1

[0048] Step 13: Calculate the frequency corresponding to the maximum power of the fast Fourier transform segment by segment:

[0049] f m (n) = maxp(fabs(F m (n))) * Fs / N, where Fs is the signal sampling frequency;

[0050] Step 14: After the calculation, obtain the instantaneous frequency curve f(m).

[0051] The instantaneous frequency characteristics of the sinusoidal frequency modulation in-pulse modulation signal are as Figure 2 shown.

[0052] Step 2: Calculate the bandwidth B of the signal through the instantaneous frequency curve. The calculation formula is as follows

[0053] B = max(f m (n)) - min(f m (n))

[0054] According to the pulse width τ of the signal, calculate the time-bandwidth product τB of the signal.

[0055] Step 3: If τB is less than the first threshold, for example, 1.5, it is an unmodulated signal or a small-bandwidth frequency modulation signal; otherwise, it is a frequency modulation signal. Proceed to Step 8.

[0056] Step 4: For unmodulated signals or small-bandwidth frequency modulation signals, use the phase difference method to calculate the instantaneous frequency. The detailed steps are as follows.

[0057] Step 41: Perform orthogonal transformation on the intermediate-frequency sampled signal S(n). The calculation formula is as follows:

[0058] S'(n) = hilbert(S(n))

[0059] Step 42: Calculate the phase of the signal using the orthogonal signal.

[0060] θ(n) = atan2(imag(S'(n)), real(S'(n)))

[0061] Step 43: Use the phase difference to find the instantaneous frequency of the signal:

[0062] f(m) = Fs(θ(n) - θ(n - 1)) / (2π), where Fs is the signal sampling frequency.

[0063] Proceed to Step 5.

[0064] Step 5: Use the instantaneous frequency obtained by the phase difference method to find the signal bandwidth.

[0065] B = max(f(m)) - min(f(m))

[0066] Perform linear fitting on the instantaneous frequency curve and calculate the error after fitting.

[0067] Step 6: If the signal bandwidth is greater than the second threshold, for example, 0.8 MHz, proceed to Step 7; otherwise, it is judged as an unmodulated signal.

[0068] Step 7: If the error after fitting is greater than the third threshold, for example, 0.7 MHz, it is judged as a non-linear frequency modulation signal; otherwise, it is judged as a linear frequency modulation signal.

[0069] Step 8: Perform linear fitting on the instantaneous frequency curve obtained by the short-time Fourier transform and calculate the error after fitting; proceed to Step 9.

[0070] Step 9: If the error after fitting is greater than the fourth threshold, for example, 0.7 MHz, it is judged as a non-linear frequency modulation signal; otherwise, it is judged as a linear frequency modulation signal.

[0071] In Step 1 of this application, the short-time Fourier transform method is used to calculate the instantaneous frequency of the intermediate-frequency sampled signal S(n), which has the characteristics of strong robustness and good performance under low signal-to-noise ratio.

[0072] In Step 4, the phase difference method is used to calculate the instantaneous frequency of the intermediate-frequency sampled signal S(n), which has the characteristics of high frequency resolution and can resolve frequency modulation signals with small bandwidths.

[0073] Through steps one to nine, the present application combines the advantages of two instantaneous frequency calculation methods to implement a method for identifying the in-pulse modulation type of a frequency-modulated signal with high robustness.

[0074] The present application combines the advantages of two instantaneous frequency modulations by leveraging the strong anti-noise ability of the short-time Fourier transform method and the high frequency resolution of the phase difference calculation method, thereby implementing a method for identifying the in-pulse modulation type of a frequency-modulated signal with strong robustness.

[0075] The embodiments of the present application described above do not constitute a limitation on the protection scope of the present application.

Claims

1. A method for identifying the in - pulse modulation type of a frequency - modulated signal by combining short - time Fourier transform and phase - difference instantaneous frequency, characterized in that, The method includes: Step 1: Calculate the instantaneous frequency of the intermediate-frequency sampled signal S(n) using the short-time Fourier transform method; Step 2: Calculate the bandwidth B of the signal through the instantaneous frequency curve; Step 3: If τB is less than the set first threshold, it is an unmodulated signal or a small-bandwidth frequency-modulated signal; otherwise, it is a frequency-modulated signal, and go to Step 8; Step 4: For an unmodulated signal or a small-bandwidth frequency-modulated signal, calculate the instantaneous frequency using the phase difference method; Step 5: Use the phase difference to obtain the instantaneous frequency of the signal and calculate the signal bandwidth; Step 6: If the signal bandwidth is greater than the set second threshold, go to Step 7; otherwise, it is determined as an unmodulated signal; Step 7: If the error after fitting is greater than the set third threshold, it is determined as a non-linear frequency-modulated signal; otherwise, it is determined as a linear frequency-modulated signal; Step 8: Perform a linear fit on the instantaneous frequency curve obtained by the short-time Fourier transform and calculate the error after fitting; go to Step 9; Step 9: If the error after fitting is greater than the set fourth threshold, it is determined as a non-linear frequency-modulated signal; otherwise, it is determined as a linear frequency-modulated signal.

2. The method according to claim 1, wherein Calculating the instantaneous frequency of the intermediate-frequency sampled signal S(n) using the short-time Fourier transform method includes: Step 11, divide the intermediate frequency sampling signal into N / 2 overlapping signal segments denoted as S m (n); where N is the number of points of the short-time Fourier transform; where m is the m-th segment and n is the n-th sampling point; Step 12, calculate S paragraph by paragraph m (n), the fast Fourier transform, is expressed as follows: F m F(n) = fft(S m (n)), n = 0:N-1 Step 13: Calculate the frequency corresponding to the maximum power of the fast Fourier transform segment by segment: f m (n) = maxp(fabs(F m (n)))*Fs / N, where Fs is the signal sampling frequency Step 14: After the calculation is completed, obtain the instantaneous frequency curve f(m).

3. The method according to claim 1, characterized in that In Step 2, the bandwidth B is determined by the following method: B = max(f m (n)) - min(f m (n)) According to the pulse width τ of the signal, calculate the time-bandwidth product τB of the signal.

4. The method according to claim 1, characterized in that, For an unmodulated signal or a small-bandwidth frequency-modulated signal, calculating the instantaneous frequency using the phase difference method includes: Step 41: Perform an orthogonal transform on the intermediate-frequency sampled signal S(n): S'(n) = hilbert(S(n)) Step 42: Calculate the phase of the signal using the orthogonal signal; θ(n) = atan2(imag(S'(n)), real(S'(n))) Step 43: Use the phase difference to obtain the instantaneous frequency of the signal: f(m) = Fs(θ(n) - θ(n - 1)) / (2π), where Fs is the signal sampling frequency.

5. The method according to claim 1, wherein In Step 5, the signal bandwidth is determined by the following method: B = max(f(m)) - min(f(m)) Perform a linear fit on the instantaneous frequency curve and calculate the error after fitting.

Images