Signal identification method based on characteristic parameters

Abstract

The invention relates to the technical field of signal analysis, and discloses a signal identification method based on characteristic parameters, comprising the following steps: S1, acquiring a frequency domain signal; s2, feature extraction; s3, obtaining a feature set by using an algorithm; s4, feature classification; and S5, during modulation mode identification output and characteristic parameter extraction, according to a modulation principle, it can be known that the instantaneous amplitude, phase and frequency of the signal carry modulation information, so that the instantaneous information of the signal can be extracted as a basis for constructing the characteristic parameters. And the characteristic set enters the classifier, and various modulation modes of the signals are output after passing through the classifier, so that the problems of low signal identification automation degree and low accuracy in a communication radio frequency signal receiving algorithm are solved.

Description

Technical Field

[0001] The present invention relates to the technical field of signal analysis, and in particular to a signal recognition method based on characteristic parameters. Background Art

[0002] Signal identification refers to the process of comparing the characteristic parameters of the measured radiation source signal with the known radiation source signal to determine the properties, type and purpose of the measured radiation source.

[0003] Automatic modulation identification technology is used in the fields of communication and radio frequency signal analysis, and is a key technology in signal analysis. The radio signal obtained by the receiver receiving antenna is converted into various modulated frequency domain signals through Fourier transform. By applying automatic modulation identification technology, the modulation mode, rate, spectrum and other characteristics of the frequency domain signal can be extracted and analyzed. The existing communication radio frequency signal receiving algorithm has the problems of low signal recognition automation and low accuracy. Therefore, a signal recognition method based on characteristic parameters is proposed. Summary of the invention

[0004] The purpose of the present invention is to provide a signal recognition method based on characteristic parameters to solve the problems raised in the above background technology.

[0005] To achieve the above object, the present invention provides the following technical solution: a signal recognition method based on characteristic parameters, comprising the following steps:

[0006] S1, obtain frequency domain signal;

[0007] S2, feature extraction;

[0008] S3, use the algorithm to obtain the feature set;

[0009] S4, feature classification;

[0010] S5, modulation mode identification output.

[0011] Preferably, in S2, when extracting characteristic parameters, according to the modulation principle, the instantaneous amplitude, phase and frequency of the signal carry the modulation information, so the instantaneous information of the signal can be extracted as the basis for constructing the characteristic parameters;

[0012] The calculation formula of the instantaneous amplitude of the signal A(n) is:

[0013]

[0014] The calculation formula of the instantaneous phase φ(n) of the signal is:

[0015]

[0016] The phase calculated according to the above formula has phase aliasing of ±π. In order to calculate the instantaneous frequency of the signal, phase aliasing must be removed to obtain the correct instantaneous frequency.

[0017]

[0018]

[0019] Preferably, in S3, the algorithm uses complex feature parameters to extract feature sets, and the feature sets extract feature parameters for signal classification based on the differences in time domain, frequency domain and instantaneous features of different types of modulation styles;

[0020] The following sub-steps are included:

[0021] S301, the maximum value γ of the power spectrum density of the normalized instantaneous amplitude max ;

[0022] S302, peak value of signal power spectrum;

[0023] S303, normalizing the deviation of the absolute value of the instantaneous amplitude;

[0024] S304, power spectrum coefficient;

[0025] S305, normalizing the peak value of the instantaneous frequency;

[0026] S306, the absolute value standard deviation of the instantaneous frequency of the zero-center non-weak signal segment;

[0027] S307. Characteristic parameters of QAM signal.

[0028] Preferably, the maximum value γ of the power spectrum density of the normalized instantaneous amplitude in S301 above max The calculation formula is:

[0029] γ max =max|DFT(a cn (i))| 2 / N s

[0030] Among them, N s is the number of sampling points, a cn (i) is the value of the zero-centered normalized instantaneous amplitude, defined as: a cn (i) = a n (i)-1, while a n (i) = a(i) / m a , is the average value of the instantaneous amplitude a(i).

[0031] Preferably, the peak value of the signal power spectrum in S302 is:

[0032] The power spectrum of the signal directly reflects the power distribution of the frequency components in the modulated signal. There will be a spectrum peak at the carrier frequency in the modulated signal. Therefore, the type of modulated signal can be distinguished according to the number of spectrum peaks in the intermediate frequency signal power spectrum and the peak state of the spectrum peak. The peak state of the spectrum peak characterizes the sharpness of the spectrum peak. The calculation formula is:

[0033]

[0034] Where p() is the square power spectrum and the fourth power spectrum of the modulated signal, and i is the subscript corresponding to the maximum value, which can be used to calculate the peak state of the square spectrum and the peak state of the fourth spectrum of the signal respectively.

[0035] Preferably, the calculation formula for the deviation of the normalized instantaneous amplitude absolute value in S303 is:

[0036]

[0037] Preferably, the power spectrum coefficient calculation formula in the above S304 is:

[0038]

[0039] p is the power spectrum of the signal, and the parameter Fpara can reflect the flatness of the signal spectrum to a certain extent.

[0040] Preferably, the peak value calculation formula of the normalized instantaneous frequency in S305 is:

[0041]

[0042] f N (n) is the instantaneous frequency normalized to zero center and is defined as: f N (n) = f n (n)-f mean ,and is the instantaneous frequency f n The average value of (i).

[0043] Preferably, the absolute value standard deviation calculation formula of the instantaneous frequency of the zero-center non-weak signal segment in S306 is:

[0044]

[0045] In the formula, a t It is an amplitude judgment threshold level for judging weak signal segments. C is the value of the threshold level when all the sampled data N s The number of non-weak signal values ​​in .

[0046] Preferably, the characteristic parameters of the QAM signal in S307 are:

[0047] Perform synchronization processing on QAM modulated signals and extract characteristic parameters from the synchronized signals;

[0048] The calculation formula for the number of values ​​greater than 0.5 in the fourth power spectrum of the modulated signal is:

[0049] N1=|P(i)| / max|P|>0.5

[0050] Where P = fft(y 4 );

[0051] The calculation formula for the number of instantaneous amplitudes greater than the mean is:

[0052] N2=|y(i)|>mean(|y|).

[0053] Compared with the prior art, the present invention adopts the above technical solution and has the following technical effects:

[0054] The present invention obtains various feature sets of the signal through a feature extraction algorithm by using the obtained frequency domain signal. The feature sets enter a classifier, and various modulation modes of the output signal after passing through the classifier, thereby solving the problems of low signal recognition automation and low accuracy in the communication radio frequency signal receiving algorithm. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0056] Figure 1 It is a schematic diagram of the process of the present invention;

[0057] Figure 2 It is a schematic diagram of the modulation identification process of the present invention;

[0058] Figure 3 is the parameter γ of the present invention max The recognition effect diagram of the modulation signal;

[0059] Figure 4 The square spectrum and the fourth power spectrum of the MPSK and 16QAM signals of the present invention;

[0060] Figure 5 It is a diagram showing the recognition effect of the spectral peak state quantity of the parameter square spectrum and the recognition effect of the quartic spectrum peak state quantity on the signal of the present invention;

[0061] Figure 6 The σ of the present inventionaa The signal recognition effect diagram;

[0062] Figure 7 This is a diagram showing the recognition effect of Fpara on a signal of the present invention;

[0063] Figure 8 For the present invention The signal recognition effect diagram;

[0064] Fig. 9 is the parameter σ of the present invention af Signal recognition effect diagram. DETAILED DESCRIPTION

[0065] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0066] Example

[0067] See also Figure 1-9 The present invention provides a technical solution: a signal recognition method based on characteristic parameters, comprising the following steps:

[0068] S1, obtain frequency domain signal; S2, feature extraction; when extracting feature parameters, according to the modulation principle, the instantaneous amplitude, phase and frequency of the signal carry the modulation information, so the instantaneous information of the signal can be extracted as the basis for constructing the feature parameters;

[0069] The calculation formula of the instantaneous amplitude of the signal A(n) is:

[0070]

[0071] The calculation formula of the instantaneous phase φ(n) of the signal is:

[0072]

[0073] The phase calculated according to the above formula has phase aliasing of ±π. In order to calculate the instantaneous frequency of the signal, phase aliasing must be removed to obtain the correct instantaneous frequency.

[0074]

[0075]

[0076] S3. Use an algorithm to obtain a feature set; the algorithm uses complex feature parameters to extract the feature set, and the feature set extracts feature parameters for signal classification based on the differences in time domain, frequency domain and instantaneous features of different types of modulation styles;

[0077] In S301, the maximum value of the power spectrum density of the normalized instantaneous amplitude γ max The calculation formula is:

[0078] γ max =max|DFT(a cn (i))| 2 / N s

[0079] Among them, N s is the number of sampling points, a cn (i) is the value of the zero-centered normalized instantaneous amplitude, defined as: a cn (i) = a n (i)-1, while a n (i) = a(i) / m a , is the average value of the instantaneous amplitude a(i);

[0080] γ max The calculation formula is, which is mainly used to distinguish signals with amplitude information from signals without amplitude recognition, γ max In analog modulation, it is used to distinguish between FM signals and AM signals. Because the instantaneous amplitude of the FM signal is constant, its zero-centered normalized instantaneous amplitude a cn (i) = 0, and its corresponding spectrum density is also zero. However, the instantaneous amplitude of the AM signal is variable, so its zero-centered normalized instantaneous amplitude a cn (i)≠0;

[0081] γ max In digital modulation, it is used to distinguish between MFSK signals and MPSK and MQAM signals. Due to the limitation of channel bandwidth, MPSK signals will have amplitude changes at the moment of phase change, so they also contain amplitude change information. The instantaneous amplitude of MFSK signals is constant, so the zero-center normalized instantaneous amplitude is zero.

[0082] like Figure 3 The data in is the parameter γ max The recognition effect diagram of the modulated signal.

[0083] In S302, the peak value of the signal power spectrum is:

[0084] The power spectrum of the signal directly reflects the power distribution of the frequency components in the modulated signal. There will be a spectrum peak at the carrier frequency in the modulated signal. Therefore, the type of modulated signal can be distinguished according to the number of spectrum peaks in the intermediate frequency signal power spectrum and the peak state of the spectrum peak. The peak state of the spectrum peak characterizes the sharpness of the spectrum peak. The calculation formula is:

[0085]

[0086] Where p(□) is the square power spectrum and the fourth power spectrum of the modulated signal, and i is the subscript corresponding to the maximum value, which can be used to calculate the peak value of the square spectrum and the peak value of the fourth power spectrum of the signal respectively;

[0087] After the MPSK signal is squared, the BPSK signal will have a large impact component at the 2-fold carrier frequency, while other MPSK and 16QAM signals with M>2 have no obvious impact components. Similarly, after the MPSK signal is squared, the QPSK signal will have a large impact component at the 4-fold carrier frequency, while other MPSK and 16QAM signals with M>4 have no obvious impact components. Therefore, the square spectrum peak state and the fourth power spectrum peak state of the signal can distinguish BPSK, QPSK and 8PSK, MQAM signals;

[0088] like Figure 4 These are the square and fourth power spectra of MPSK and 16QAM signals.

[0089] like Figure 5 The diagram shows the recognition effect of the spectral peak state of the parameter square spectrum on the signal and the recognition effect of the spectral peak state of the fourth power spectrum on the signal.

[0090] In S303, the calculation formula for the deviation of the normalized instantaneous amplitude absolute value is:

[0091]

[0092] The deviation of the normalized instantaneous amplitude absolute value can be used to distinguish between signals containing absolute direct amplitude signals and signals that do not contain absolute direct amplitude signals. The instantaneous amplitude of the MQAM signal has different amplitudes depending on M and has absolute direct amplitude information, while the 8PSK signal itself does not contain an amplitude signal and can be used to distinguish between MQAM and 8PSK signals.

[0093] like Figure 6 is the parameter σ aa Signal recognition effect diagram.

[0094] In S304, the power spectrum coefficient calculation formula is:

[0095]

[0096] p is the power spectrum of the signal. The parameter Fpara can reflect the flatness of the signal spectrum to a certain extent. If the power spectrum is dominated by obvious impact components, the ratio will be very large.

[0097] like Figure 7 This is the recognition effect diagram of the parameter Fpara on the signal.

[0098] In S305, the peak value calculation formula of the normalized instantaneous frequency is:

[0099]

[0100] f N (n) is the instantaneous frequency normalized to zero center and is defined as: f N (n) = f n (n)-f mean ,and is the instantaneous frequency f n (i) the average value;

[0101] Different digital modulation signals have different frequency distributions. The parameters that reflect the instantaneous frequency distribution of the signal are Can classify MFSK signals and MPSK / QAM signals;

[0102] like Figure 8 For parameters Signal recognition effect diagram.

[0103] In S306, the absolute value standard deviation of the instantaneous frequency of the zero-center non-weak signal segment is calculated as:

[0104]

[0105] In the formula, a t It is an amplitude judgment threshold level for judging weak signal segments. C is the value of the threshold level when all the sampled data N s The number of non-weak signal values ​​in ;

[0106] like Fig. 9 is the parameter σ af Signal recognition effect diagram.

[0107] In S307, characteristic parameters of QAM signal:

[0108] Perform synchronization processing on QAM modulated signals and extract characteristic parameters from the synchronized signals;

[0109] The calculation formula for the number of values ​​greater than 0.5 in the fourth power spectrum of the modulated signal is:

[0110] N1=|P(i)| / max|P|>0.5

[0111] Where P = fft(y 4 );

[0112] The calculation formula for the number of instantaneous amplitudes greater than the mean is:

[0113] N2=|y(i)|>mean(|y|)

[0114] S4. Feature classification: Feature classification based on ADMRA algorithm classifier:

[0115] In S3, an algorithm is used to obtain a feature set. The definition and expression characteristics of feature parameters can be used in classification algorithms. Different algorithms can be generated according to the order in which the feature parameters are used. The modulation recognition algorithm (ADMRA) is essentially to select different thresholds for each feature parameter to maximize the recognition rate. Since M signal segments may have different classification results, a large number of judgments must still be made after multiple judgments.

[0116] The following focuses on the threshold determination of the ADMRA algorithm:

[0117] In order to realize the recognition of analog and digital modulation, the threshold values ​​of seven characteristic parameters are defined: Ku2, Ku4, Fpara, Garma_max, Sigma_aa, Sigma_af, Mu42f. At the same time, a normalized amplitude threshold is also defined. The determination of the threshold values ​​of the characteristic parameters is based on the above-mentioned large number of algorithms and experiments. According to the definition of the characteristic parameters and the value of the threshold values ​​of each characteristic parameter, the flow chart of the ADMRA algorithm can be obtained. Figure 2 ;

[0118] S5, modulation mode identification output: according to different modulation signal types, select the characteristic parameters of the signal and formulate a modulation mode identification scheme;

[0119] Among them, the characteristic parameters used in the modulation recognition process are as follows:

[0120] ①The peak value of Ku2 square spectrum;

[0121] ②The peak value of Ku4 quartic spectrum;

[0122] ③Fpara power spectrum coefficient;

[0123] ④Garma_max is the maximum value of the power spectrum density of the normalized instantaneous amplitude;

[0124] ⑤Sigma_aa is the deviation of the absolute value of the normalized instantaneous amplitude;

[0125] ⑥Sigma_af is the standard deviation of the absolute value of the instantaneous frequency of the non-weak signal segment;

[0126] ⑦The peak value of the normalized instantaneous frequency of Mu42f.

[0127] In summary, through the obtained frequency domain signal, various feature sets of the signal are obtained through the feature extraction algorithm, the feature set enters the classifier, and various modulation methods of the signal are output after the classifier, which solves the problems of low signal recognition automation and low accuracy in the communication RF signal receiving algorithm.

[0128] So far, the embodiments of the present invention have been described in detail in conjunction with the accompanying drawings. It should be noted that the implementation methods not shown or described in the drawings or the body of the specification are all forms known to ordinary technicians in the relevant technical field and are not described in detail. In addition, the above definitions of each component are not limited to the various specific structures, shapes or methods mentioned in the embodiments, and ordinary technicians in the field can simply change or replace them.

[0129] It will be appreciated by those skilled in the art that the features described in the various embodiments and / or claims of the present invention may be combined and / or combined in various ways, even if such combinations and / or combinations are not explicitly described in the present invention. In particular, the features described in the various embodiments and / or claims of the present invention may be combined and / or combined in various ways without departing from the spirit and teachings of the present invention. All of these combinations and / or combinations fall within the scope of the present invention.

[0130] The specific embodiments described above further illustrate the objectives, technical solutions and beneficial effects of the present invention in detail. It should be understood that the above description is only a specific embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A signal recognition method based on characteristic parameters, characterized in that: The steps include: S1, obtain frequency domain signal; S2, feature extraction; S3, use the algorithm to obtain the feature set; S4, feature classification; S5. Modulation mode identification output.

2. A signal recognition method based on characteristic parameters according to claim 1, characterized in that: In S2, when extracting characteristic parameters, according to the modulation principle, the instantaneous amplitude, phase and frequency of the signal carry the modulation information. Therefore, the instantaneous information of the signal can be extracted as the basis for constructing characteristic parameters. The calculation formula of the instantaneous amplitude of the signal A(n) is: The calculation formula of the instantaneous phase φ(n) of the signal is: The phase calculated according to the above formula has phase aliasing of ±π. In order to calculate the instantaneous frequency of the signal, phase aliasing must be removed to obtain the correct instantaneous frequency.

3. The signal recognition method based on characteristic parameters according to claim 1, characterized in that: In S3, the algorithm uses complex feature parameters to extract feature sets. The feature sets extract feature parameters for signal classification based on the differences in time domain, frequency domain, and instantaneous features of different types of modulation styles. The following sub-steps are included: S301, the maximum value γ of the power spectrum density of the normalized instantaneous amplitude max ; S302, peak value of signal power spectrum; S303, normalizing the deviation of the absolute value of the instantaneous amplitude; S304, power spectrum coefficient; S305, normalizing the peak value of the instantaneous frequency; S306, the absolute value standard deviation of the instantaneous frequency of the zero-center non-weak signal segment; S307. Characteristic parameters of QAM signal.

4. The signal recognition method based on characteristic parameters according to claim 3 is characterized in that: The maximum value γ of the power spectrum density of the normalized instantaneous amplitude in S301 max The calculation formula is: γ max =max|DFT(a cn (i))| 2 *N s Among them, N s is the number of sampling points, a cn (i) is the value of the zero-centered normalized instantaneous amplitude, defined as: a cn (i) = a n (i)-1, while a n (i) = a(i) / m a , is the average value of the instantaneous amplitude a(i).

5. The signal recognition method based on characteristic parameters according to claim 3 is characterized in that: The peak value of the signal power spectrum in S302 is: The power spectrum of the signal directly reflects the power distribution of the frequency components in the modulated signal. There will be a spectrum peak at the carrier frequency in the modulated signal. Therefore, the type of modulated signal can be distinguished according to the number of spectrum peaks in the intermediate frequency signal power spectrum and the peak state of the spectrum peak. The peak state of the spectrum peak characterizes the sharpness of the spectrum peak. The calculation formula is: in are the square power spectrum and the fourth power spectrum of the modulated signal, i is the subscript corresponding to the maximum value, and the peak state of the square spectrum and the peak state of the fourth spectrum of the signal can be calculated respectively.

6. The signal recognition method based on characteristic parameters according to claim 3 is characterized in that: The calculation formula of the deviation of the normalized instantaneous amplitude absolute value in S303 is:

7. The signal recognition method based on characteristic parameters according to claim 3 is characterized in that: The power spectrum coefficient calculation formula in S304 is: p is the power spectrum of the signal, and the parameter Fpara can reflect the flatness of the signal spectrum to a certain extent.

8. The signal recognition method based on characteristic parameters according to claim 3 is characterized in that: The peak value calculation formula of the normalized instantaneous frequency in S305 is: f N (n) is the instantaneous frequency normalized to zero center and is defined as: f N (n) = f n (n)-f mean ,and is the instantaneous frequency f n The average value of (i).

9. The signal recognition method based on characteristic parameters according to claim 3, characterized in that: The calculation formula for the absolute value standard deviation of the instantaneous frequency of the zero-center non-weak signal segment in S306 is: In the formula, a t It is an amplitude judgment threshold level for judging weak signal segments. C is the value of the threshold level when all the sampled data N s The number of non-weak signal values ​​in .

10. The signal recognition method based on characteristic parameters according to claim 3, characterized in that: The characteristic parameters of the QAM signal in S307 are: Perform synchronization processing on QAM modulated signals and extract characteristic parameters from the synchronized signals; The calculation formula for the number of values ​​greater than 0.5 in the fourth power spectrum of the modulated signal is: N1=|P(i)| / max|P|>0.5 Where P = fft(y 4 ); The calculation formula for the number of instantaneous amplitudes greater than the mean is: N2=|y(i)|>mean(|y|).

Images