Robustness conventional radar signal identification method under complex conditions

Abstract

The invention belongs to the technical field of radar radiation source signal intra-pulse analysis, and particularly relates to a robust conventional radar signal identification method under complex conditions. According to the method, direct recognition of conventional radar signals is realized through the steps of signal time-frequency structure characteristic analysis, adaptive parameter setting, adaptive preprocessing, multi-feature-domain characteristic measurement, multi-index joint judgment and the like, the radar signal intra-pulse processing flow is greatly shortened, and the method is not influenced by acquisition conditions, signal parameters, random interference and the like; the method can adapt to actual complex application scenes, and has the advantages of low calculation complexity, high robustness and the like.

Description

Technical Field

[0001] The invention belongs to the technical field of intra-pulse analysis of radar radiation source signals, and in particular relates to a robust conventional radar signal recognition method under complex conditions. Background Art

[0002] With the continuous development of radar technology, new systems and new purpose radars are constantly put into use, and radar signal waveform modulation systems are also constantly improving and perfecting. The corresponding radar modulation types are increasing, the signal form is becoming more complex, and the modulation parameters are also becoming more agile and diverse. Under this circumstance, the difficulty of signal recognition is increasing, especially the recognition of radar signal intra-pulse modulation types faces greater challenges. Traditional modulation recognition algorithms can only recognize limited signal types and cannot adapt to the increasingly diverse signal types and increasingly complex signal environments. They cannot meet the needs of robust and accurate recognition under complex conditions such as low signal-to-noise ratio caused by non-ideal factors in the transmission and reception links, the presence of interference signals, changes in transmission conditions, and changes in signal parameters.

[0003] In the past, the idea of ​​identifying the intra-pulse modulation of radar signals was to calculate the value of one or more index parameters, and then make judgments from a number of candidate types in turn according to the preset rules and the parameter conditions. For example, the number of signal spectrum peaks is calculated to distinguish between conventional radar signals and frequency shift keying FSK signals. If it is an FSK signal, it is further determined which type of FSK it is.

[0004] There are many existing radar modulation recognition methods, and commonly used methods include time domain analysis, frequency domain analysis, time-frequency analysis, and high-order statistical methods (Tang Yan. Radar signal parameter detection and intra-pulse recognition [D]. University of Electronic Science and Technology of China, 2021, Liu Fangjian, Li He, Chen Yunfan, et al. Algorithm simulation and verification based on radar signal intra-pulse modulation recognition [J]. Journal of Qingdao University (Engineering Technology Edition) 2023.38(2):68–74.). The time domain analysis method mainly refers to the instantaneous autocorrelation method, which usually uses the instantaneous phase and instantaneous frequency curves and the corresponding autocorrelation. This type of method requires the extraction of the instantaneous amplitude, alias-free instantaneous phase and instantaneous frequency of the signal. The instantaneous phase and instantaneous frequency are more sensitive to noise, so this method has a high signal-to-noise ratio requirement. The frequency domain analysis method includes spectrum analysis, cepstrum analysis, etc. This type of method mainly refers to distinguishing different types of intra-pulse modulation by analyzing the spectrum, power spectrum or cepstrum of the signal. It focuses on the overall situation of different frequency components in the signal, and does not have the ability to analyze local characteristics. Many signals have no obvious differences in the frequency domain. Therefore, the ability of this method to distinguish is very limited. Time-frequency analysis methods mainly include wavelet transform, short-time Fourier transform, etc. The wavelet transform method uses wavelet transform for multi-scale analysis, has a certain ability to remove noise, and can detect the phase jump point of the phase signal, but has no obvious advantage in the recognition ability of other types of signals. Therefore, it is difficult to use for accurate recognition of all types, and the algorithm has high computational complexity and long processing time. In the time-frequency analysis method, the short-time Fourier transform method has a large amount of calculation and there is a contradiction between time and frequency accuracy; Wigner-Ville analysis has serious cross terms and is mostly used to detect linear or curved modulated signals. High-order statistical methods mainly refer to the use of parameters such as high-order moment accumulation to identify signal types. This type of method has good signal-to-noise ratio performance, but the amount of calculation is large, and some signal parameters require to be known.

[0005] Among many signal types, conventional radar signals (also known as NS signals) have the simplest structure, obvious characteristics, and only contain a single frequency component. They are the most widely used in practice. When analyzing the intrapulse modulation type of radar signals, if conventional radar signals can be accurately identified at the first time, processing time can be greatly saved and the processing flow can be shortened, that is, conventional radar signals can be directly confirmed.

[0006] Ideally, the instantaneous frequency of the NS signal is a smooth straight line. However, in actual applications, the phase and frequency will change to a certain extent after being affected by noise and interference. The instantaneous frequency has random fluctuations and glitches, making it difficult to confirm directly. In the past, it was difficult to directly confirm by simply counting the peaks in the signal frequency domain, which would lead to misjudgment as other types in the subsequent process and increase the processing time (Li Mingyan, Zhang Luyun, Jiang Mingyan, et al. Recognition method of complex intrapulse modulated radar signal [J]. Computer Engineering and Applications, 2011, 47(15): 5. DOI: 10.3778 / j.issn.1002-8331.2011.15.042.").

[0007] Therefore, in order to obtain accurate signal information and quickly and accurately determine the modulation type, especially to identify conventional radar signals at the first time, it is necessary to carry out targeted processing and identification based on the robust modulation characteristics of conventional radar signals. Summary of the invention

[0008] In order to realize fast and accurate recognition of conventional radar signals in actual application scenarios, the present invention combines the structural characteristics of conventional radar signals in the time and frequency domains to design a robust conventional radar signal recognition method under complex conditions. Direct recognition of conventional radar signals is achieved through steps such as signal time-frequency structural characteristic analysis, adaptive parameter setting, adaptive preprocessing, multi-feature domain characteristic measurement, and multi-index joint judgment, which greatly shortens the radar signal pulse processing process and is not affected by acquisition conditions, signal parameters, random interference, etc. It can adapt to actual complex application scenarios and has the advantages of low computational complexity and high robustness.

[0009] The technical solution adopted by the present invention is: a robust conventional radar signal recognition method under complex conditions, which is divided into the following steps:

[0010] S1. Signal reception

[0011] The received signal to be processed is represented by s0(n0), where n0 represents the index value of the sampling point, n0 = 1, 2, ..., N0, N0 represents the initial signal length, and the signal sampling rate is f s ;

[0012] S2. Adaptive narrowband filtering

[0013] S2.1 uses the real sinusoidal signal frequency estimation method based on DFT to estimate the peak frequency point position f0 of the signal s0(n0) (for specific methods, see the document "[4] Guo Luxiang. Research on real sinusoidal signal frequency estimation algorithm based on DFT [D], Southeast University, 2019.").

[0014] S2.2 estimates the bandwidth BW0 of the signal s0(n0), the specific steps are as follows:

[0015] S2.2.1 According to the non-parametric power spectrum estimation method, the power spectrum S(f) of the signal s0(n0) is obtained, where f represents the frequency index value of the power spectrum, and f = -f s / 2,...,f s / 2.

[0016] S2.2.2 Find the peak position f of S(f) max , and from f max Search to both sides and record the frequency value f when S(f) drops to 3dB of the peak value l and f r , f l Indicates the frequency value 3dB to the left of the peak, f r Indicates the frequency value 3dB to the right of the peak.

[0017] S2.2.3 Calculation of f l and f r The distance between them is obtained, and the bandwidth estimation value BW0 of the signal s0(n0) is obtained. r -f l .

[0018] S2.3 Adaptively filter the signal using an FFT filter with f0 as the center frequency according to the estimated peak frequency position f0 and bandwidth BW0. The specific steps are as follows:

[0019] S2.3.1 Set the filter bandwidth factor δ and set the starting frequency of the frequency domain filter to f start =f0-δ·BW0, the cut-off frequency is f end =f0+δ·BW0. Since the conventional radar NS signal is a single-frequency narrow-band signal, δ is generally set to 0.55-0.85 in practical applications.

[0020] S2.3.2 Perform N on s0(n0) FFT Fourier transform of the point, convert it to the frequency domain, and then keep f start and f end The frequency components between the two, all other frequency components are set to zero, and finally an inverse Fourier transform is performed to obtain the filtered signal s1(n0), usually N FFT Take 1024.

[0021] S3. Pulse leading and trailing edge detection

[0022] S3.1 Detect the envelope starting point n of the filtered signal s1(n0) start and end point n end , the specific method is as follows:

[0023] S3.1.1 Calculate the signal envelope a(n0) = |s1(n0)| and calculate the average value of the envelope

[0024] S3.1.2 Setting the signal detection threshold th A , the detection signal envelope exceeds the threshold a(n0)≥th A The starting point index value n corresponding to part A start and the end point index n end . Generally take th A =0.5.

[0025] S3.2 Using n start and n end Intercept the leading and trailing edges of signal s1(n0), that is, take the signal index value n0=n start ,...,n end The corresponding s1(n0) part is taken as the intercepted signal s(n), where n represents the new signal index, n=1,2,...,N, N=n end -n start +1 indicates the new signal length after truncation.

[0026] S4. Pulse width determination

[0027] S4.1 Calculate the pulse width PW of the intercepted signal s(n) according to the following formula:

[0028] PW=N / f s ;

[0029] S4.2 In actual situations, short pulses are short signals and generally do not contain complex modulation, and are usually conventional radar signals. Therefore, they can be judged according to the pulse width PW: if PW is less than 0.5μs, the signal is a short pulse and is directly judged as a conventional radar signal, ending the entire processing flow; if PW is greater than 0.5μs, proceed to step S5.

[0030] S5. Instantaneous frequency calculation

[0031] S5.1 In order to take into account both the detailed information on the instantaneous frequency and the robustness under complex conditions, the differential order L is adaptively set according to the signal length N. According to engineering experience, L can generally be set to N / 30;

[0032] S5.2 estimates the instantaneous frequency of the signal s(n) according to the L-fold phase difference method. The specific calculation steps are as follows:

[0033] S5.2.1 Calculate the signal amplitude:

[0034]

[0035] Among them, s I(n), s Q (n) respectively represent the real and imaginary parts of the signal s(n): s(n) = s I (n)+i×s Q (n), i represents the imaginary unit, and Imag(·) represents the operation of taking an imaginary number.

[0036] S5.2.2 Use L-fold phase difference method to find the instantaneous frequency sequence p(n) to improve the adaptability to noise:

[0037]

[0038] S6. Characteristic curve fitting

[0039] S6.1 with straight line Fit p(n) and use the least squares estimation method to get the fitting slope Fitting bias value and the fitted instantaneous frequency series For the least squares estimation method, see “[6]Steven M.Kay, Luo Pengfei. Fundamentals of Statistical Signal Processing[M]. Publishing House of Electronics Industry, 2014.”

[0040] S6.2 Calculate the characteristic curve fitting error ε:

[0041]

[0042] S6.3 Determine the error situation: If the error ε is less than the threshold th ε , then proceed to step S8 to calculate the subsequent judgment index; otherwise, proceed to step S7 for time-frequency analysis to update the fitting result. ε It is generally set to 0.015~0.1.

[0043] S7. Time-frequency analysis

[0044] Perform time-frequency analysis on the signal s(n), and update the instantaneous frequency sequence p(n) based on the time-frequency analysis results to further reduce the influence of random factors such as noise interference and improve the robustness of the algorithm. The specific steps are as follows:

[0045] S7.1 Perform short-time Fourier transform on the signal s(n) to obtain the time-frequency matrix in Represents N×N FFT dimensional complex set matrix;

[0046] S7.2 Calculate the energy of the time-frequency diagram matrix Μ0(n,f), that is, find the amplitude value point by point to obtain the time-frequency energy: Μ(n,f)=|Μ0(n,f)|.

[0047] S7.3 Calculate the mean of the time-frequency energy M(n,f):

[0048] S7.4 extracts the time-frequency distribution. The specific steps are as follows:

[0049] S7.4.1 Setting the time-frequency energy threshold th M , generally set to 0.7~0.9;

[0050] S7.4.2 Compare M(n,f) at time n with The size of n will satisfy The total f-value is expressed as Indicates that at time n, The kth frequency value of , k = 1, 2, ..., K, K is the total number of frequencies that meet the conditions;

[0051] S7.4.3 Calculate the value of each column The mean of :

[0052] S7.5 Fit the curve q(n) according to the method in step S6 to obtain the corresponding fitting slope Fitting bias value and the fitted instantaneous frequency series Then the fitting error ε1 is obtained;

[0053] S7.6 compares the size of ε1 and ε: if ε1 < ε, update the fitting result in step S6, update p(n) to q(n), and Updated to Will Updated to Update ε to ε1, and then proceed to step S8; otherwise, proceed directly to step S8.

[0054] S8. Dual Linear Frequency Modulation Signal Identification

[0055] Perform dual-line frequency modulation signal discrimination to avoid misjudging conventional radar signals as dual-line frequency modulation signals; the specific steps are as follows:

[0056] S8.1 performs summing and smoothing on the instantaneous frequency sequence p(n). The smoothing order is λ, usually λ = 50, and the smoothed sequence g(n) is calculated as follows:

[0057]

[0058] S8.2 Find the minimum and maximum values ​​of g(n), and replace the index corresponding to the maximum value with n max Indicates that the index corresponding to the minimum value is n min express;

[0059] S8.3 determines whether the signal is an upper V-shaped dual linear frequency modulation signal. The specific process is as follows:

[0060] S8.3.1 Calculate the flag F1 as follows:

[0061]

[0062] Among them, ∧ represents the "AND" operation, and ∨ represents the "OR" operation.

[0063] If F1 is not zero, that is, the signal preliminarily meets the characteristics of the upper V-shaped dual linear frequency modulation signal, then proceed to S8.3.2; if F1 is zero, then proceed to S8.4.

[0064] S8.3.2 V-type dual linear frequency modulation signal discrimination:

[0065] S8.3.2.1 Compute the first-order difference of g(n):

[0066] S8.3.2.2 Take n=2,...,n min -1 corresponds to Part, calculation The number of elements less than 0 in

[0067] S8.3.2.3 Take n=n min ,...,N corresponds to Part, calculation The number of elements greater than 0 in

[0068] S8.3.2.4 Setting the bilinear threshold th N ,calculate and The bilinear threshold is usually set to th N =0.65;

[0069] S8.3.2.5 If Then the signal is identified as an upper V-shaped bilinear frequency modulation signal, and the bilinear flag F is set DLFM is 1, then proceed to step S9; if Then proceed to step S8.4.

[0070] S8.4 determines whether the signal is a lower V-shaped dual linear frequency modulation signal:

[0071] S8.4.1 Calculate the flag F2 as follows:

[0072]

[0073] If F2 is not zero, that is, the signal initially meets the characteristics of the V-type bilinear frequency modulation signal, then proceed to S8.4.2; if F1 is zero, then set the bilinear flag F DLFM =0, proceed to step S9.

[0074] S8.4.2 V-type dual linear frequency modulation signal discrimination.

[0075] S8.4.2.1 If not Then calculate the first-order difference of g(n): (Already calculated in step S8.3.2.1 However, depending on the data, it is possible that step S8.3.2.1 has not been entered before, so it is necessary to calculate it here); if it already exists Then

[0076] S8.4.2.2 Take n=2,...,n max -1 corresponds to Part, calculation The number of elements greater than 0 in

[0077] S8.4.2.3 Take n=n max ,...,N corresponds to Part, calculation The number of elements less than 0 in

[0078] S8.4.2.4 Setting the bilinear threshold th N ,calculate and The bilinear threshold is usually set to th N =0.65;

[0079] S8.4.2.5 If Then the signal is identified as a lower V-shaped bilinear frequency modulation signal, and the bilinear flag F is set DLFM is 1; if Then F DLFM is 0.

[0080] S9. Instantaneous frequency variance characteristics

[0081] Calculate the variance characteristics of the instantaneous frequency to prevent misjudgment with phase modulated signals or frequency modulated signals. The specific steps are as follows:

[0082] S9.1 Calculate the mean of the instantaneous frequency sequence p(n) Then calculate the variance σ1 of the instantaneous frequency sequence p(n):

[0083]

[0084] S9.2 Combine the instantaneous frequency series fitted in step S6.1 Calculate the linear variance σ2 of the instantaneous frequency p(n):

[0085]

[0086] S9.3 Calculate the absolute value LV of the difference between the variance σ1 and the linear variance σ2:

[0087] LV = |σ1-σ2|;

[0088] S10. Time-Bandwidth Product Calculation

[0089] S10.1 estimate the bandwidth BW of the signal s(n) according to the method in step S2.2;

[0090] S10.2 combines the pulse width PW in step S4.1 and the bandwidth BW in step S10.1 to calculate the time-bandwidth product TB of the signal:

[0091] TB=PW·BW;

[0092] S11. Conventional radar signal judgment

[0093] S11.1 Calculate the adaptive modulation slope threshold th of signal s(n) K :

[0094]

[0095] S11.2 Calculate the flag F3 according to the following formula:

[0096]

[0097] If F3 is not zero, the signal is determined to be a conventional radar signal; otherwise, the signal is another type of signal.

[0098] Technical effect:

[0099] The present invention proposes a robust conventional radar signal recognition method under complex conditions to adapt to the accurate recognition of conventional radar signals in actual non-ideal application scenarios. Specifically, it includes the design of a processing flow for conventional radar signal recognition, adaptive signal processing and adaptive parameter setting for complex transmission and reception conditions, and multiple judgment indicators and calculation methods based on the time domain and frequency domain structural characteristics of conventional radar signals, which shortens the processing flow of the existing radar intra-pulse analysis system and improves the accuracy and robustness of radar modulation type recognition.

[0100] The technical benefit of the present invention is to directly identify and confirm the type of conventional radar signals under complex conditions, and to support the rapid and accurate identification of radar targets by the radar intra-pulse analysis system in complex actual scenarios. This is mainly reflected in:

[0101] (1) A processing flow for directly identifying conventional radar signals under complex conditions is designed. The modulation slope estimation value, characteristic curve fitting error, instantaneous frequency variance characteristics, time-width-bandwidth product and other indicators are combined for comprehensive judgment. In order to further shorten the processing flow, the pulse width judgment is first performed after the detection and processing in combination with the actual medium and short pulse situation, which can directly judge the conventional radar signal and save the subsequent calculation link.

[0102] (2) The present invention designs a method for obtaining the instantaneous frequency domain curve of the signal based on time-frequency analysis, and uses the signal distribution in the time-frequency image to obtain the main trend of the instantaneous frequency of the signal to increase the robustness against noise interference and other situations;

[0103] (3) In order to avoid misjudging it as a dual-linear frequency modulation signal, a discrimination method for two different types of dual-linear frequency modulation signals is designed;

[0104] (4) To avoid misjudging the signals as FSK or PSK modulation signals, the variance characteristics of the instantaneous frequency are used for differentiation;

[0105] (5) Robust adaptive signal processing and adaptive parameter setting are specially performed for actual complex scenarios, including signal adaptive bandwidth filtering, leading and trailing edge detection, and time-frequency analysis combined with error conditions. In terms of threshold setting, adaptive threshold values ​​are set based on signal conditions, such as the signal modulation slope threshold th K . BRIEF DESCRIPTION OF THE DRAWINGS

[0106] Figure 1 It is a flow chart of the present invention;

[0107] Figure 2 is a flow chart of step S7 time-frequency analysis;

[0108] Figure 3 is a flow chart of step S8 dual linear frequency modulation signal discrimination;

[0109] Figure 4 It is the time domain waveform of an ideal conventional radar signal;

[0110] Figure 5 It is the spectrum waveform of an ideal conventional radar signal;

[0111] Figure 6 is the instantaneous frequency of the ideal conventional radar signal obtained in step S5;

[0112] Figure 7It is the time-frequency image of an ideal conventional radar signal;

[0113] Figure 8 It is the time domain waveform of the conventional radar signal when the signal-to-noise ratio is 0dB;

[0114] Fig. 9 It is the frequency domain waveform of the conventional radar signal when the signal-to-noise ratio is 0dB;

[0115] Fig.10 It is the first-order phase difference of the conventional radar signal when the signal-to-noise ratio is 0dB;

[0116] Fig.11 is the instantaneous frequency waveform of the conventional radar signal obtained in step S5 when the signal-to-noise ratio is 0 dB;

[0117] Fig.12 It is the time-frequency diagram of the conventional radar signal when the signal-to-noise ratio is 0dB;

[0118] Fig.13 is the time-frequency distribution of the conventional radar signal obtained in step S7.4 when the signal-to-noise ratio is 0 dB;

[0119] Fig.14 It is the instantaneous frequency q(n) of the conventional radar signal obtained in step S7 when the signal-to-noise ratio is 0 dB. DETAILED DESCRIPTION

[0120] The specific implementation modes of the present invention will be further described below in conjunction with the accompanying drawings.

[0121] Figure 1 : is a flow chart of the present invention. The present invention proposes a robust conventional radar signal recognition method under complex conditions, which is divided into the following steps:

[0122] S1, signal reception, mainly refers to obtaining the signal to be analyzed;

[0123] S2, adaptive narrowband filtering, mainly estimates the frequency and bandwidth of the signal, and adaptively sets the filtering parameters according to the signal frequency and bandwidth to complete the signal filtering;

[0124] S3, pulse leading and trailing edge detection, mainly performs pulse detection on the signal obtained after filtering and intercepts the signal part;

[0125] S4, pulse width judgment, calculates the pulse width of the signal, and first identifies the short-pulse conventional radar signal according to the pulse width;

[0126] S5, instantaneous frequency estimation, mainly calculates the instantaneous frequency of the signal based on the phase difference method;

[0127] S6, characteristic curve fitting, mainly fits the instantaneous frequency of the signal based on the least squares estimation method to obtain the fitting slope, fitting error, etc.;

[0128] S7, time-frequency analysis, mainly performs time-frequency analysis on the signal with an error higher than the threshold in step S6, and updates the instantaneous frequency and its corresponding least square fitting result in combination with the characteristics of the time-frequency graph to improve the robustness of the algorithm;

[0129] S8, bilinear modulation signal discrimination, mainly to avoid misjudgment as bilinear frequency modulation signal, and calculate corresponding signs for two different types of bilinear frequency modulation signals;

[0130] S9, instantaneous frequency variance characteristic, mainly calculates the linear variance of instantaneous frequency to avoid misjudgment with phase modulation and frequency modulation signals;

[0131] S10, time-bandwidth product calculation, mainly estimates the signal bandwidth and calculates the time-bandwidth product to distinguish it from the pulse compression signal;

[0132] S11, conventional radar signal judgment, mainly confirms the signal type based on the indicator parameters of the previous multiple domains to determine whether it is a conventional radar signal.

[0133] Figure 2 This is the time-frequency analysis flow chart of step S7. In this step, the signal with a large fitting error in the previous characteristic curve fitting process is further processed to increase the robustness to noise interference and other situations, that is, the corresponding instantaneous frequency characteristics are extracted using the time-frequency image. The specific process is to perform a short-time Fourier transform on the signal, calculate the representative frequency value at each moment based on the time-frequency matrix, combine to form a new instantaneous frequency, and then fit the characteristic curve again, and compare the two fitting errors. If the error corresponding to the time-frequency analysis is small, the fitting instantaneous frequency, fitting parameters and fitting error obtained at this time are used in the subsequent steps.

[0134] Figure 3 This is the flow chart of step S8 dual-linear frequency modulation signal determination. This step is mainly to determine whether the signal is a dual-linear frequency modulation signal. The determination method adopted is to first observe whether the instantaneous frequency of the signal is generally in an upper V shape or a lower V shape (step S8.3.1 and step S8.4.1). After basically meeting the requirements, further determine whether the two parts of the dual-linear frequency modulation corresponding to the signal are the main body of the signal (step S8.3.2 and step S8.4.2). If the requirements are met, the corresponding label F DLFM =1.

[0135] Figure 4 It is the time domain waveform of an ideal conventional radar signal. This figure shows the image of the signal amplitude changing with time in the time domain of the simulated conventional radar signal s0(n0) under ideal conditions. The horizontal axis in the figure is time in seconds; the vertical axis is the amplitude value of the signal. There is a zero point before and after the signal.

[0136] Figure 5 This is the spectrum waveform of an ideal conventional radar signal. This figure shows the image of an ideal conventional radar signal s0(n0) in the frequency domain. It can be seen that there is a narrow-band single peak in the figure, that is, the conventional radar signal has a simple frequency domain structure and only one frequency component.

[0137] Figure 6 is the instantaneous frequency of the ideal conventional radar signal obtained by step S5. The horizontal axis represents the sample point index of the signal, and the vertical axis is the frequency value. The figure shows the instantaneous frequency curve image of the ideal conventional radar signal calculated by the L-phase difference method in step S5. It should be noted that since the signal is an ideal signal, the filtering process of step S2 is not required, and the leading and trailing edge detection of step S3 and step S4 are not performed. Instead, the result obtained by step S5 is directly performed on the received signal. It can be seen that the middle part of the curve is a horizontal straight line, which corresponds to the signal part, that is, the signal has only a fixed frequency component and does not change with time. The ups and downs at both ends of the curve are caused by the continuous zero points before and after the signal.

[0138] Figure 7 It is the time-frequency image obtained by short-time Fourier transform of the ideal conventional radar signal. The horizontal axis is time in seconds, and the vertical axis is frequency value in Hz. It can be seen that the signal is a horizontal straight line in the time-frequency diagram. It should be noted that because the signal is an ideal signal, no adaptive narrowband filtering is performed. This figure is the original received signal directly after step S3 and S7.2 to obtain the short-time Fourier transform corresponding to M (n, f).

[0139] Figure 8 This is the time domain waveform of a conventional radar signal when the signal-to-noise ratio is 0dB. The conventional radar signal in this figure is accompanied by additive white Gaussian noise, and the signal-to-noise ratio is 0dB. Figure 4 Compared with the time domain waveform of the ideal signal in the figure, it can be seen that the signal edge changes greatly, and the signal and the front and rear zero point boundaries are not obvious, indicating that there is little difference between the signal and the electromagnetic environment noise in the time domain.

[0140] Fig. 9 This is the frequency domain waveform of a conventional radar signal when the signal-to-noise ratio is 0dB. This figure shows the spectrum image of a conventional radar signal when the signal-to-noise ratio is 0dB. It can be seen that the signal has only a single frequency point. Figure 5 Compared with the ideal signal in the image, the noise floor level of the signal is significantly increased. However, compared with the time domain waveform, the signal position can be clearly located in the frequency domain.

[0141] Fig.10is the first-order phase difference of a conventional radar signal when the signal-to-noise ratio is 0dB. This figure is the instantaneous frequency curve of a conventional radar signal when the signal-to-noise ratio is 0dB obtained using the first-order phase difference method commonly used in the past. The horizontal axis represents the index value of the signal sample point, and the vertical axis represents the normalized signal frequency value. It can be seen that Figure 6 In comparison, due to the influence of noise, the curve in the figure no longer has the obvious characteristics of a conventional radar signal, and it is difficult to directly determine the signal type based on this result.

[0142] Fig.11 It is the instantaneous frequency waveform of the conventional radar signal obtained by step S5 when the signal-to-noise ratio is 0dB. The figure shows the instantaneous frequency curve image of the conventional radar signal when the signal-to-noise ratio is 0dB, calculated by the L-phase difference method in step S5. The horizontal axis represents the sample point index of the signal, and the vertical axis is the frequency value. It should be noted that the signal has not been filtered in step S2, nor has the leading and trailing edge detection of step S3 and step S4 been performed, but the result obtained by directly performing step S5 on the received signal. It can be seen that the middle part of the curve is horizontal in overall trend, corresponding to the signal part, but there are large fluctuations, and it cannot be directly interpreted as a single fixed frequency component, so it cannot be directly determined as a conventional radar signal.

[0143] Fig.12 This is the time-frequency diagram of a conventional radar signal when the signal-to-noise ratio is 0dB. The horizontal axis of the diagram is time in seconds, and the vertical axis is frequency in Hz. It can be seen that due to the high noise level, compared to Figure 7 There are a lot of snowflake-like noise points in the ideal signal in the figure, but there is still an obvious horizontal straight line in the time-frequency diagram, which shows that the time-frequency analysis method has good robustness. It should be noted that this figure does not perform adaptive narrowband filtering on the signal, but the original received signal is directly obtained by short-time Fourier transform in steps S3 and S7.2 to obtain M(n,f).

[0144] Fig.13 is the time-frequency distribution of the conventional radar signal obtained by step S7.4 when the signal-to-noise ratio is 0dB. Specifically, the figure is a signal with a signal-to-noise ratio of 0dB directly proceeding to step S7 after steps S2 and S3. The corresponding value in the figure is the value that satisfies step S7.4.2 All frequency points of . It can be seen that compared with Fig.12 There are snowflakes caused by noise in Fig.13 Only the frequency part corresponding to the signal is retained. The instantaneous frequency extracted on this basis can better suppress noise, thereby achieving a robust effect.

[0145] Fig.14 is the instantaneous frequency q(n) of the conventional radar signal obtained in step S7 when the signal-to-noise ratio is 0dB. Specifically, the figure is Fig.13The instantaneous frequency q(n) is obtained based on the result after step S7.4.3. It can be seen that compared with Fig.10 and Fig.11 The instantaneous frequency image in is Fig.14 There is only one fixed frequency component in the signal, which can be directly determined as a conventional radar signal with obvious characteristics, indicating that this method has strong signal feature extraction capability and robustness.

Claims

1. A robust conventional radar signal recognition method under complex conditions, characterized in that: The method is divided into the following steps: S1. Signal reception The received signal to be processed is represented by s0(n0), where n0 represents the index value of the sampling point, n0 = 1, 2, ..., N0, N0 represents the initial signal length, and the signal sampling rate is f s ; S2. Adaptive narrowband filtering S2.1 uses the real sinusoidal signal frequency estimation method based on DFT to estimate the peak frequency position f0 of the signal s0(n0); S2.2 estimates the bandwidth BW0 of the signal s0(n0), the specific steps are as follows: S2.2.1 According to the non-parametric power spectrum estimation method, the power spectrum S(f) of the signal s0(n0) is obtained, where f represents the frequency index value of the power spectrum, and f = -f s / 2,...,f s / 2; S2.2.2 Find the peak position f of S(f) max , and from f max Search to both sides and record the frequency value f when S(f) drops to 3dB of the peak value l and f r , f l Indicates the frequency value 3dB to the left of the peak, f r Indicates the frequency value 3dB to the right of the peak; S2.2.3 Calculation of f l and f r The distance between them is obtained, and the bandwidth estimation value BW0 of the signal s0(n0) is obtained. r -f l ; S2.3 Adaptively filter the signal using an FFT filter with f0 as the center frequency according to the estimated peak frequency position f0 and bandwidth BW0. The specific steps are as follows: S2.3.1 Set the filter bandwidth factor δ and set the starting frequency of the frequency domain filter to f start =f0-δ·BW0, the cut-off frequency is f end =f0+δ·BW0; S2.3.2 Perform N on s0(n0) FFT Fourier transform of the point, convert it to the frequency domain, and then keep f start and f end The frequency components between , all other frequency components are set to zero, and finally an inverse Fourier transform is performed to obtain the filtered signal s1(n0); S3. Pulse leading and trailing edge detection S3.1 Detect the envelope starting point n of the filtered signal s1(n0) start and end point n end , the specific method is as follows: S3.1.1 Calculate the signal envelope a(n0) = |s1(n0)| and calculate the average value of the envelope S3.1.2 Setting the signal detection threshold th A , the detection signal envelope exceeds the threshold a(n0)≥th A The starting point index value n corresponding to part A start and the end point index n end ; S3.2 Using n start and n end Intercept the leading and trailing edges of signal s1(n0), that is, take the signal index value n0=n start ,...,n end The corresponding s1(n0) part is taken as the intercepted signal s(n), where n represents the new signal index, n=1,2,...,N, N=n end -n start +1 indicates the new signal length after interception; S4. Pulse width determination S4.1 Calculate the pulse width PW of the intercepted signal s(n) according to the following formula: PW=N / f s ; S4.2: Determine according to the pulse width PW: if PW is less than 0.5 μs, the signal is a short pulse and is directly determined to be a conventional radar signal, and the entire processing flow ends; if PW is greater than 0.5 μs, proceed to step S5; S5. Instantaneous frequency calculation S5.1 adaptively sets the differential order L according to the signal length N; S5.2 estimates the instantaneous frequency of the signal s(n) according to the L-fold phase difference method. The specific calculation steps are as follows: S5.2.1 Calculate the signal amplitude: Among them, s I (n), s Q (n) respectively represent the real and imaginary parts of the signal s(n): s(n) = s I (n)+i×s Q (n), i represents the imaginary unit, and Imag(·) represents the operation of taking imaginary numbers; S5.2.2 Use L-fold phase difference method to find the instantaneous frequency sequence p(n) to improve the adaptability to noise: S6. Characteristic curve fitting S6.1 with straight line Fit p(n) and use the least squares estimation method to get the fitting slope Fitting bias value and the fitted instantaneous frequency series S6.2 Calculate the characteristic curve fitting error ε: S6.3 Determine the error situation: If the error ε is less than the threshold th ε , then proceed to step S8 to calculate the subsequent judgment index; otherwise, perform the time-frequency analysis of step S7 to update the fitting result; S7. Time-frequency analysis Perform time-frequency analysis on the signal s(n), and update the instantaneous frequency sequence p(n) based on the time-frequency analysis results to further reduce the influence of random factors such as noise interference and improve the robustness of the algorithm. The specific steps are as follows: S7.1 Perform short-time Fourier transform on the signal s(n) to obtain the time-frequency matrix in Represents N×N FFT dimensional complex set matrix; S7.2 Calculate the energy of the time-frequency diagram matrix Μ0(n,f), that is, calculate the amplitude value point by point to obtain the time-frequency energy: Μ(n,f)=|Μ0(n,f)|; S7.3 Calculate the mean of the time-frequency energy M(n,f): S7.4 extracts the time-frequency distribution. The specific steps are as follows: S7.4.1 Setting the time-frequency energy threshold th M ; S7.4.2 Compare M(n,f) at time n with The size of n will satisfy The total f-value is expressed as Indicates that at time n, The kth frequency value of , k = 1, 2, ..., K, K is the total number of frequencies that meet the conditions; S7.4.3 Calculate the value of each column The mean of : S7.5 According to the method in step S6, the curve q(n) is fitted to obtain the corresponding fitting slope Fitting bias value and the fitted instantaneous frequency series Then the fitting error ε1 is obtained; S7.6 compares the size of ε1 and ε: if ε1 < ε, update the fitting result in step S6, update p(n) to q(n), and Updated to Will Updated to Update ε to ε1, and then proceed to step S8; otherwise, proceed directly to step S8; S8. Dual Linear Frequency Modulation Signal Identification Perform dual-line frequency modulation signal discrimination to avoid misjudging conventional radar signals as dual-line frequency modulation signals; the specific steps are as follows: S8.1 performs summing and smoothing on the instantaneous frequency sequence p(n): the smoothing order is λ, and the smoothed sequence g(n) is calculated as follows: S8.2 Find the minimum and maximum values ​​of g(n), and replace the index corresponding to the maximum value with n max Indicates that the index corresponding to the minimum value is n min express; S8.3 determines whether the signal is an upper V-shaped dual linear frequency modulation signal. The specific process is as follows: S8.3.1 Calculate the flag F1 as follows: Among them, ∧ represents the "and" operation, and ∨ represents the "or" operation; If F1 is not zero, that is, the signal preliminarily meets the characteristics of the upper V-shaped dual linear frequency modulation signal, then proceed to S8.3.2; if F1 is zero, then proceed to S8.4; S8.3.2 V-type dual linear frequency modulation signal discrimination: S8.3.2.1 Compute the first-order difference of g(n): S8.3.2.2 Take n=2,...,n min -1 corresponds to Part, calculation The number of elements less than 0 in S8.3.2.3 Take n=n min ,...,N corresponds to Part, calculation The number of elements greater than 0 in S8.3.2.4 Setting the bilinear threshold th N ,calculate and of and; S8.3.2.5 If Then the signal is identified as an upper V-shaped bilinear frequency modulation signal, and the bilinear flag F is set DLFM is 1, then proceed to step S9; if Then proceed to step S8.4; S8.4 determines whether the signal is a lower V-shaped dual linear frequency modulation signal: S8.4.1 Calculate the flag F2 as follows: If F2 is not zero, that is, the signal initially meets the characteristics of the V-type bilinear frequency modulation signal, then proceed to S8.4.2; if F1 is zero, then set the bilinear flag F DLFM =0, proceed to step S9; S8.4.2 V-type dual linear frequency modulation signal discrimination: S8.4.2.1 If not Then calculate the first-order difference of g(n): If already exists Then S8.4.2.2 Take n=2,...,n max -1 corresponds to Part, calculation The number of elements greater than 0 in S8.4.2.3 Take n=n max ,...,N corresponds to Part, calculation The number of elements less than 0 in S8.4.2.4 Setting the bilinear threshold th N ,calculate and of and; S8.4.2.5 If Then the signal is identified as a lower V-shaped bilinear frequency modulation signal, and the bilinear flag F is set DLFM is 1; if Then F DLFM is 0; S9. Instantaneous frequency variance characteristics Calculate the variance characteristics of the instantaneous frequency to prevent misjudgment with phase modulated signals or frequency modulated signals; the specific steps are as follows: S9.1 Calculate the mean of the instantaneous frequency sequence p(n) Then calculate the variance σ1 of the instantaneous frequency sequence p(n): S9.2 Combine the instantaneous frequency series fitted in step S6.1 Calculate the linear variance σ2 of the instantaneous frequency p(n): S9.3 Calculate the absolute value LV of the difference between the variance σ1 and the linear variance σ2: LV = |σ1-σ2|; S10. Time-Bandwidth Product Calculation S10.1 estimate the bandwidth BW of the signal s(n) according to the method in step S2.2; S10.2 combines the pulse width PW in step S4.1 and the bandwidth BW in step S10.1 to calculate the time-bandwidth product TB of the signal: TB=PW·BW; S11. Conventional radar signal judgment S11.1 Calculate the adaptive modulation slope threshold th of signal s(n) K : S11.2 Calculate the flag F3 according to the following formula: If F3 is not zero, the signal is determined to be a conventional radar signal; otherwise, the signal is another type of signal.

2. A robust conventional radar signal recognition method under complex conditions according to claim 1, characterized in that: In S2.3.1, the filter bandwidth factor δ is set to 0.55-0.

85.

3. A robust conventional radar signal recognition method under complex conditions according to claim 1, characterized in that: S2.3.2 N FFT Take 1024.

4. A robust conventional radar signal recognition method under complex conditions according to claim 1, characterized in that: In S3.1.2, the signal detection threshold th A =0.

5.

5. A robust conventional radar signal recognition method under complex conditions according to claim 1, characterized in that: In S5.1, the difference order L=N / 30.

6. A robust conventional radar signal recognition method under complex conditions according to claim 1, characterized in that: In S6.3, the threshold th ε Set to 0.015~0.

1.

7. A robust conventional radar signal recognition method under complex conditions according to claim 1, characterized in that: In S7.4.1, the time-frequency energy threshold is set to 0.7 to 0.

9.

8. A robust conventional radar signal recognition method under complex conditions according to claim 1, characterized in that: In S8.1, the smoothing order λ=50.

9. A robust conventional radar signal recognition method under complex conditions according to claim 1, characterized in that: In S8.3.2.4, the bilinear threshold th N =0.

65.

10. A robust conventional radar signal recognition method under complex conditions according to claim 1, characterized in that: In S8.4.2.4, the bilinear threshold th N =0.65.

Images