Blind identification method, device and system for signal modulation mode

Abstract

The invention relates to a blind recognition method, device and system for a signal modulation mode, and the method comprises the steps: building a target recognition model which is used for recognizing a signal modulation mode corresponding to a received signal; when a signal is received, extracting each instantaneous feature of the signal to obtain an instantaneous feature set corresponding to the signal; for each instantaneous feature in the instantaneous feature set, determining an instantaneous feature statistic corresponding to the signal according to the instantaneous feature so as to obtain an instantaneous feature statistic set corresponding to the signal; and inputting the instantaneous characteristic statistic set into a target recognition model, so that the target recognition model outputs a target signal modulation mode corresponding to the signal. According to the invention, the accuracy and reliability of blind identification of the signal modulation mode corresponding to the signal are improved.

Description

Technical Field

[0001] The present application relates to the field of signal processing technology, and in particular to a method, device and system for blindly identifying a signal modulation mode. Background Art

[0002] In a traditional communication system, when a receiving end receives a signal, since the receiving end knows the signal modulation mode of the signal, the receiving end can demodulate the received signal according to the known signal modulation mode to restore the original information. However, with the rapid development of wireless communication technology, various wireless communication systems are widely used. In many practical scenarios, the receiving end is not clear about the signal modulation mode corresponding to the received signal. Therefore, the receiving end needs to analyze the received signal to achieve blind recognition of the signal modulation mode corresponding to the received signal, and then use the recognized signal modulation mode to demodulate the received signal.

[0003] However, when the receiving end blindly identifies the signal modulation mode corresponding to the received signal, it usually analyzes the time domain characteristics or frequency domain characteristics of the received signal to achieve blind identification of the signal modulation mode. However, the above blind identification method uses a single signal feature to blindly identify the signal, which makes the blind identification of the signal inaccurate and difficult to meet the requirements of high reliability. Summary of the invention

[0004] The present application provides a method, device and system for blind recognition of signal modulation modes to solve the problems of inaccurate recognition and low recognition reliability in blind recognition of received signals in the prior art.

[0005] In a first aspect, the present application provides a method for blindly identifying a signal modulation mode, comprising:

[0006] Establishing a target recognition model, wherein the target recognition model is used to identify a signal modulation mode corresponding to a received signal;

[0007] When a signal is received, each instantaneous feature of the signal is extracted to obtain an instantaneous feature set corresponding to the signal;

[0008] For each of the instantaneous features in the instantaneous feature set, determine an instantaneous feature statistic corresponding to the signal according to the instantaneous feature, so as to obtain an instantaneous feature statistic set corresponding to the signal;

[0009] The instantaneous feature statistic set is input into the target recognition model so that the target recognition model outputs a target signal modulation mode corresponding to the signal.

[0010] In an optional embodiment, when the instantaneous feature set includes instantaneous amplitude, the instantaneous feature statistic corresponding to the signal includes a first feature statistic, and the first feature statistic represents the maximum value of the instantaneous amplitude spectral density;

[0011] Determining, according to the instantaneous feature, an instantaneous feature statistic corresponding to the signal includes:

[0012] When the instantaneous feature set includes the instantaneous amplitude, obtaining the number of samples corresponding to the signal;

[0013] Performing zero-center normalization processing on the instantaneous amplitude to obtain the processed instantaneous amplitude;

[0014] The processed instantaneous amplitude and the number of samples are input into a first feature determination formula to obtain the first feature statistic, wherein the first feature determination formula includes:

[0015] γ max =maxDFT(A cn (n)) 2 N

[0016] In the above formula, γ max represents the first characteristic statistic, max represents the maximum value, DFT() represents discrete Fourier transform, A cn (n) represents the instantaneous amplitude after processing, and N represents the number of samples.

[0017] In an optional embodiment, when the instantaneous feature set includes the instantaneous amplitude, the instantaneous feature statistics corresponding to the signal also include a second feature statistic, a third feature statistic and a fourth feature statistic, wherein the second feature statistic represents the standard deviation of the zero-centered normalized non-weak signal segment instantaneous amplitude, the third feature statistic represents the standard deviation of the absolute value of the zero-centered normalized instantaneous amplitude, and the fourth feature statistic represents the compactness of the zero-centered normalized instantaneous amplitude;

[0018] Determining, according to the instantaneous feature, an instantaneous feature statistic corresponding to the signal includes:

[0019] Determine the first number that is not a weak signal from the signals of the sampled number;

[0020] Inputting the first number and the processed instantaneous amplitude into a second feature determination formula to obtain the second feature statistic;

[0021] Inputting the number of samples and the processed instantaneous amplitude into a third feature determination formula to obtain the third feature statistic;

[0022] The processed instantaneous amplitude is input into the fourth feature determination formula to obtain the fourth feature statistic; wherein,

[0023] The second feature determination formula includes:

[0024]

[0025] The third characteristic determination formula includes:

[0026]

[0027] The fourth characteristic determination formula includes:

[0028]

[0029] In the above formula, σ da represents the second characteristic statistic, A cn (n) represents the instantaneous amplitude after processing, C represents the first number, σ aa represents the third characteristic statistic, N represents the number of samples, represents the fourth characteristic statistic, A cn (n)>at represents a non-weak signal, at represents a threshold level, and E() represents a Gaussian density function.

[0030] In an optional embodiment, when the instantaneous feature set includes an instantaneous phase, the instantaneous feature statistics corresponding to the signal include a fifth feature statistic and a sixth feature statistic, the fifth feature statistic characterizing the standard deviation of the instantaneous phase nonlinear component of the zero-center non-weak signal segment, and the sixth feature statistic characterizing the standard deviation of the absolute value of the instantaneous phase nonlinear component of the zero-center non-weak signal segment;

[0031] Determining, according to the instantaneous feature, an instantaneous feature statistic corresponding to the signal includes:

[0032] When the instantaneous feature set includes the instantaneous phase, performing zero center processing on the instantaneous phase to obtain a nonlinear component of the instantaneous phase;

[0033] The first number and the nonlinear component are respectively input into the fifth feature determination formula and the sixth feature determination formula to obtain the fifth feature statistic and the sixth feature statistic; wherein,

[0034] The fifth characteristic determination formula includes:

[0035]

[0036] The sixth characteristic determination formula includes:

[0037]

[0038] In the above formula, σ dp represents the fifth characteristic statistic, φ NL (n) represents the nonlinear component, C represents the first number, σ ap represents the sixth characteristic statistic, A cn (n)>at indicates a non-weak signal, and at indicates a threshold level.

[0039] In an optional embodiment, when the instantaneous feature set includes an instantaneous frequency, the instantaneous feature statistics corresponding to the signal include a seventh feature statistic and an eighth feature statistic, the seventh feature statistic characterizing the standard deviation of the absolute value of the instantaneous frequency of the zero-centered normalized non-weak signal segment, and the eighth feature statistic characterizing the compactness of the zero-centered normalized instantaneous frequency;

[0040] Determining, according to the instantaneous feature, an instantaneous feature statistic corresponding to the signal includes:

[0041] When the instantaneous feature set includes the instantaneous frequency, the instantaneous frequency is processed to obtain a first instantaneous frequency and a second instantaneous frequency, wherein the first instantaneous frequency represents a zero-centered normalized non-weak signal segment instantaneous frequency, and the second instantaneous frequency represents a zero-centered normalized instantaneous frequency;

[0042] Inputting the first number and the first instantaneous frequency into a seventh feature determination formula to obtain the seventh feature statistic;

[0043] The second instantaneous frequency is input into the eighth feature determination formula to obtain the eighth feature statistic; wherein,

[0044] The seventh characteristic determination formula includes:

[0045]

[0046] The eighth characteristic determination formula includes:

[0047]

[0048] In the above formula, σ af represents the seventh characteristic statistic, C represents the first number, represents the eighth characteristic statistic, A cn (n)>at indicates a non-weak signal, at indicates the threshold level, E() indicates the Gaussian density function, f N (n) represents the first instantaneous frequency, f cn (n) represents the second instantaneous frequency.

[0049] In an optional implementation, the establishing of the target recognition model includes:

[0050] Acquire a first preset signal set and a preset signal modulation mode corresponding to each first preset signal in the first preset signal set;

[0051] Performing preset instantaneous feature extraction on each of the first preset signals in the first preset signal set to obtain a preset instantaneous feature set corresponding to each of the preset signals;

[0052] For each of the preset instantaneous features in the preset instantaneous feature set corresponding to each of the first preset signals in the first preset signal set, determining, according to the preset instantaneous features, a preset instantaneous feature statistic corresponding to the preset signal, so as to obtain a preset instantaneous feature statistic set corresponding to each of the first preset signals in the first preset signal set;

[0053] Associating a preset instantaneous feature statistic set corresponding to each of the first preset signals in the first preset signal set with a preset signal modulation mode corresponding to each of the first preset signals in the first preset signal set to obtain a first association relationship;

[0054] The preset machine learning model is trained using the first association relationship to establish the target recognition model.

[0055] In an optional implementation, when receiving a signal, extracting each instantaneous feature of the signal to obtain an instantaneous feature set corresponding to the signal includes:

[0056] Determine whether the received signal is in a second preset signal set, where a second preset signal in the second preset signal set represents a preset signal of a known signal modulation mode;

[0057] When the received signal is not in the second preset signal set, each instantaneous feature of the signal is extracted to obtain an instantaneous feature set corresponding to the signal.

[0058] In an optional embodiment, the method further includes:

[0059] Establishing a decision tree model, wherein the decision tree model is used to identify a signal modulation mode corresponding to a received signal;

[0060] Before performing the step of extracting each instantaneous feature of the signal to obtain an instantaneous feature set corresponding to the signal, the method further includes:

[0061] When the signal is received, determining whether the received signal is in a third preset signal set, wherein a third preset signal in the third preset signal set represents a preset signal whose signal modulation mode can be identified by using the decision tree model;

[0062] When the received signal is in the third preset signal set, using the decision tree model to identify the target signal modulation mode corresponding to the signal;

[0063] When the received signal is not in the third preset signal set, the step of extracting each instantaneous feature of the signal to obtain an instantaneous feature set corresponding to the signal is performed.

[0064] In a second aspect, the present application provides a blind recognition device for a signal modulation mode, comprising:

[0065] An establishment module is used to establish a target recognition model, wherein the target recognition model is used to identify a signal modulation mode corresponding to a received signal;

[0066] An extraction module, used for extracting each instantaneous feature of the signal when receiving the signal, so as to obtain an instantaneous feature set corresponding to the signal;

[0067] A determination module, configured to determine, for each of the instantaneous features in the instantaneous feature set, an instantaneous feature statistic corresponding to the signal according to the instantaneous feature, so as to obtain an instantaneous feature statistic set corresponding to the signal;

[0068] The recognition module is used to input the instantaneous feature statistic set into the target recognition model so that the target recognition model outputs the target signal modulation mode corresponding to the signal.

[0069] In a third aspect, the present application provides a blind recognition system for a signal modulation mode, including a signal sending device and a signal receiving device;

[0070] The signal sending device is used to send a signal to the signal receiving device;

[0071] The signal receiving device is used to establish a target recognition model, and when receiving a signal, executes the blind recognition method of the signal modulation method as described above. The target recognition model is used to identify the signal modulation method corresponding to the received signal.

[0072] The above-mentioned technical scheme provided by the embodiment of the present application has the following advantages compared with the prior art. The blind identification method of the signal modulation mode provided by the embodiment of the present application includes establishing a target recognition model, and the target recognition model is used to identify the signal modulation mode corresponding to the received signal; when the signal is received, each instantaneous feature of the signal is extracted to obtain an instantaneous feature set corresponding to the signal; for each instantaneous feature in the instantaneous feature set, the instantaneous feature statistic corresponding to the signal is determined according to the instantaneous feature to obtain an instantaneous feature statistic set corresponding to the signal; the instantaneous feature statistic set is input into the target recognition model so that the target recognition model outputs the target signal modulation mode corresponding to the signal. Through the above method, the present application pre-establishes a target recognition model for identifying the signal modulation method corresponding to the received signal, so that when the signal is received, the instantaneous features of the signal are extracted, and the instantaneous feature statistics corresponding to each instantaneous feature are determined by using the extracted instantaneous features, so that all the obtained instantaneous feature statistics are input into the pre-trained target recognition model, and then the target recognition model outputs the target signal modulation method corresponding to the signal, avoiding the drawbacks of blind identification of the signal modulation method using a single signal feature, and improving the accuracy and reliability of blind identification of the signal modulation method corresponding to the signal. BRIEF DESCRIPTION OF THE DRAWINGS

[0073] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the invention and, together with the description, serve to explain the principles of the invention.

[0074] In order to more clearly illustrate the embodiments of the present invention 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, for ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0075] One or more embodiments are exemplarily described by pictures in the corresponding drawings, and these exemplified descriptions do not constitute limitations on the embodiments. Elements with the same reference numerals in the drawings represent similar elements, and unless otherwise stated, the figures in the drawings do not constitute proportional limitations.

[0076] Figure 1 A flow chart of a blind identification method of a signal modulation mode provided in an embodiment of the present application;

[0077] Figure 2 A schematic diagram of a feature extraction process provided in an embodiment of the present application;

[0078] Figure 3A schematic diagram of the structure of a blind recognition system for a signal modulation mode provided in an embodiment of the present application;

[0079] Figure 4 A schematic diagram of the structure of a blind recognition device for a signal modulation mode provided in an embodiment of the present application;

[0080] Figure 5 A schematic diagram of the structure of an electronic device provided in an embodiment of the present application;

[0081] In the above attached figure:

[0082] 301. signal sending device; 302. signal receiving device;

[0083] 401, establishment module; 402, extraction module; 403, determination module; 404, identification module;

[0084] 501, processor; 502, communication interface; 503, memory; 504, communication bus. DETAILED DESCRIPTION

[0085] In order to make the purpose, technical solution and advantages of the embodiments of the present application clearer, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of this application.

[0086] The disclosure below provides many different embodiments or examples to implement different structures of the present invention. In order to simplify the disclosure of the present invention, the parts and settings of specific examples are described below. Of course, they are only examples, and the purpose is not to limit the present invention. In addition, the present invention can repeat reference numbers and / or letters in different examples. This repetition is for the purpose of simplification and clarity, and does not itself indicate the relationship between the various embodiments and / or settings discussed.

[0087] refer to Figure 1 , Figure 1 A flow chart of a blind identification method of a signal modulation mode provided in an embodiment of the present application. A blind identification method of a signal modulation mode provided in this embodiment comprises the following steps:

[0088] S101: Establish a target recognition model.

[0089] In this embodiment, the target recognition model is used to identify the signal modulation mode corresponding to the received signal. The target recognition model is actually a trained machine learning model. Based on the classification and recognition ability of the machine learning model, it can obtain higher computational complexity and detection performance, and is suitable for high-precision signal recognition under unfavorable factors such as strong interference and fast-changing harsh electromagnetic environment and lack of prior information. Therefore, in this embodiment, a target recognition model is established so that when a signal is received, the received signal can be blindly recognized according to the established target recognition model to improve the accuracy and reliability of signal blind recognition.

[0090] In the above, the target recognition model is established, including:

[0091] Acquire a first preset signal set and a preset signal modulation mode corresponding to each first preset signal in the first preset signal set;

[0092] Extracting preset instantaneous features from each first preset signal in the first preset signal set to obtain a preset instantaneous feature set corresponding to each preset signal;

[0093] For each preset instantaneous feature in the preset instantaneous feature set corresponding to each first preset signal in the first preset signal set, determine the preset instantaneous feature statistic corresponding to the preset signal according to the preset instantaneous feature, so as to obtain the preset instantaneous feature statistic set corresponding to each first preset signal in the first preset signal set;

[0094] Associating a preset instantaneous feature set corresponding to each first preset signal in the first preset signal set with a preset signal modulation mode corresponding to each first preset signal in the first preset signal set to obtain a first association relationship;

[0095] The preset machine learning model is trained using the first association relationship to establish a target recognition model.

[0096] Among them, the first preset signal set may include single-carrier signals of different modulations, for example, BPSK, 8PSK, QAM, 16QAM and 64QAM, etc. The preset signal modulation mode corresponding to each first preset signal can be set according to actual needs, and is not specifically limited in this embodiment. After obtaining each first preset signal in the first preset signal set, the preset instantaneous features such as the preset instantaneous amplitude, preset instantaneous phase and preset instantaneous frequency of each first preset signal can be extracted, so as to obtain the preset instantaneous feature set corresponding to each first preset signal. For each preset instantaneous feature in each preset instantaneous feature set, the preset instantaneous feature statistic set corresponding to the signal is determined according to the preset instantaneous feature, so as to associate each preset instantaneous feature statistic set with the preset signal modulation mode corresponding to each first preset signal, so as to train the preset machine learning model using the above-mentioned association relationship to establish a target recognition model.

[0097] Specifically, when training the preset machine learning model, the loss function is pre-set, and the trainable parameters are updated according to the gradient calculated in each training process. The learning rate determines the step of parameter update, and the optimal combination of the two is further explored to avoid overfitting or underfitting. The training samples are divided into training sets and validation sets, and the network is effectively trained through cross-validation, and the trained network is used to perform blind recognition verification of signal modulation methods.

[0098] More specifically, cross entropy is used as the loss function, and the loss function includes:

[0099]

[0100] In the above formula, L represents the loss function, M represents the number of samples, N represents the number of categories, and p ic represents the probability that the i-th sample belongs to category C, x ic ∈{0,1} represents the sample label, which is encoded using one-hot encoding.

[0101] It should be noted that the method for determining the preset instantaneous features mentioned above is consistent with the method for determining the instantaneous features described below, and the method for determining the preset instantaneous feature statistics mentioned above is consistent with the method for determining the instantaneous feature statistics described below. For details, please refer to the following description, which will not be repeated in this embodiment.

[0102] S102: When a signal is received, each instantaneous feature of the signal is extracted to obtain an instantaneous feature set corresponding to the signal.

[0103] In this embodiment, the instantaneous feature set includes instantaneous amplitude, instantaneous phase and instantaneous frequency. Assuming that the carrier frequency of the signal has been accurately estimated, the signal is frequency-converted to obtain the modulation signal expression as follows:

[0104]

[0105] In the above formula, x(t) represents the signal, a k represents the transmitted code, M represents the length of the transmitted code, e represents the signal energy coefficient, p(t) represents the transmitted code waveform, T s is the code element width, θ c is the carrier phase.

[0106] After obtaining the expression of the modulated signal, the signal can be converted into A / D and the analytical expression of the signal can be obtained through Hilbert transform, as follows:

[0107] x(n)=x s (n)+jx c (n)

[0108] In the above formula, x(n) represents the signal, x s (n) represents the real part, which is the original signal sequence signal, the actual physical quantity part, and contains the amplitude information of the signal itself. c (n) represents the imaginary part, which contains the behavior information of the signal.

[0109] The above signal analytical expression is converted to obtain the instantaneous amplitude, instantaneous phase and instantaneous frequency. The expressions of instantaneous amplitude, instantaneous phase and instantaneous frequency are as follows:

[0110] Instantaneous amplitude:

[0111] Instantaneous phase:

[0112] Instantaneous frequency:

[0113] In the above formula, A(n) represents the instantaneous amplitude, θ(n) represents the instantaneous phase, f(n) represents the instantaneous frequency, and f s Indicates the sampling frequency.

[0114] In the above step S102, when a signal is received, each instantaneous feature of the signal is extracted to obtain an instantaneous feature set corresponding to the signal, including:

[0115] Determine whether the received signal is in a second preset signal set, where a second preset signal in the second preset signal set represents a preset signal of a known signal modulation mode;

[0116] When the received signal is not in the second preset signal set, each instantaneous feature of the signal is extracted to obtain an instantaneous feature set corresponding to the signal.

[0117] In order to reduce unnecessary waste of computing resources in this embodiment, when a signal is received, it is determined whether the signal is a preset signal of a known signal modulation mode. If the signal is a preset signal of a known signal modulation mode, there is no need to call the target recognition model to blindly recognize the target signal modulation mode corresponding to the signal, and the signal can be directly demodulated using the preset signal modulation mode corresponding to the signal. If the signal is not a preset signal of a known signal modulation mode, it is necessary to extract each instantaneous feature of the signal to obtain an instantaneous feature set corresponding to the signal, thereby using the instantaneous feature set and the target recognition model to realize blind recognition of the target signal modulation mode corresponding to the signal.

[0118] It should be noted that each second preset signal in the second preset signal set can be set according to actual needs, and this embodiment does not limit each second preset signal in the second preset signal set.

[0119] In this embodiment, a blind identification method of a signal modulation mode provided in this embodiment further includes:

[0120] A decision tree model is established, and the decision tree model is used to identify the signal modulation mode corresponding to the received signal.

[0121] Before executing the step of extracting each instantaneous feature of the signal in step S102 to obtain an instantaneous feature set corresponding to the signal, the blind recognition method of a signal modulation mode provided in this embodiment further includes the following steps:

[0122] When a signal is received, determining whether the received signal is in a third preset signal set, wherein a third preset signal in the third preset signal set represents a preset signal whose signal modulation mode can be identified using a decision tree model;

[0123] When the received signal is in the third preset signal set, using a decision tree model to identify a target signal modulation mode corresponding to the signal;

[0124] When the received signal is not in the third preset signal set, the step of extracting each instantaneous feature of the signal to obtain an instantaneous feature set corresponding to the signal is performed.

[0125] Among them, the decision tree model is an existing blind recognition model of the signal modulation mode, and the specific details can be referred to the existing technology, which is not limited in this embodiment. Since it is not necessary to determine the instantaneous feature statistics according to the instantaneous features when establishing the decision tree model, the resources used by the decision tree model are less than those used by the machine learning model. Therefore, when the decision tree model can be used to blindly identify the signal modulation mode of the received signal, the decision tree model can be used directly without calling the target recognition model to reduce unnecessary waste of computing resources.

[0126] Specifically, each third preset signal can be pre-set to use the decision tree model to identify the target signal modulation mode corresponding to the signal to determine whether the received signal is in the third preset signal set. If the received signal is the third preset signal, the decision tree model can be directly used to identify the target signal modulation mode corresponding to the signal; if the received signal is not the third preset signal, it is necessary to call the target recognition model to identify the target signal modulation mode corresponding to the signal.

[0127] S103: for each instantaneous feature in the instantaneous feature set, determine the instantaneous feature statistic corresponding to the signal according to the instantaneous feature, so as to obtain the instantaneous feature statistic set corresponding to the signal.

[0128] In this embodiment, since the target recognition model is obtained by training the preset instantaneous feature statistics corresponding to the preset instantaneous features of each first preset signal, when each instantaneous feature corresponding to the signal is obtained, the instantaneous feature statistics corresponding to each instantaneous feature are determined, thereby using the determined instantaneous feature statistics to realize blind recognition of the target signal modulation mode corresponding to the signal, so as to improve the accuracy and reliability of blind recognition of the signal modulation mode.

[0129] In the above, when the instantaneous feature set includes the instantaneous amplitude, the instantaneous feature statistic corresponding to the signal includes a first feature statistic, and the first feature statistic represents the maximum value of the instantaneous amplitude spectral density. In the above step S103, the instantaneous feature statistic corresponding to the signal is determined according to the instantaneous feature, including:

[0130] When the instantaneous feature set includes the instantaneous amplitude, obtaining the number of samples corresponding to the signal;

[0131] Perform zero-center normalization processing on the instantaneous amplitude to obtain the processed instantaneous amplitude;

[0132] The processed instantaneous amplitude and the number of samples are input into a first feature determination formula to obtain a first feature statistic, wherein the first feature determination formula includes:

[0133] γ max =maxDFT(A cn (n)) 2 N

[0134] In the above formula, γ max represents the first characteristic statistic, max represents the maximum value, DFT() represents discrete Fourier transform, A cn (n) represents the instantaneous amplitude after processing, and N represents the number of samples.

[0135] Among them, the instantaneous amplitude after processing is determined by the following formula:

[0136]

[0137] In the above formula, A(n) represents the instantaneous amplitude and N represents the number of samples.

[0138] Since the maximum value of the zero-centered normalized instantaneous amplitude spectral density characterizes the instantaneous amplitude change of the signal, it reflects the changing characteristics of the signal envelope and distinguishes between constant envelope and non-constant envelope signal modulation modes.

[0139] In this embodiment, when the instantaneous feature set includes the instantaneous amplitude, the instantaneous feature statistics corresponding to the signal also include the second feature statistics, the third feature statistics and the fourth feature statistics, the second feature statistics characterizes the standard deviation of the zero-centered normalized non-weak signal segment instantaneous amplitude, the third feature statistics characterizes the standard deviation of the absolute value of the zero-centered normalized instantaneous amplitude, and the fourth feature statistics characterizes the compactness of the zero-centered normalized instantaneous amplitude. In the above step S103, the instantaneous feature statistics corresponding to the signal are determined according to the instantaneous features, including:

[0140] Determine the first number that is not a weak signal from the signal of the number of samples;

[0141] Inputting the first number and the processed instantaneous amplitude into a second feature determination formula to obtain a second feature statistic;

[0142] Inputting the number of samples and the processed instantaneous amplitude into a third feature determination formula to obtain a third feature statistic;

[0143] The processed instantaneous amplitude is input into the fourth feature determination formula to obtain the fourth feature statistic; wherein,

[0144] The second feature determination formula includes:

[0145]

[0146] The third feature determination formula includes:

[0147]

[0148] The fourth characteristic determination formula includes:

[0149]

[0150] In the above formula, σ da represents the second characteristic statistic, A cn (n) represents the instantaneous amplitude after processing, C represents the first number, σ aa represents the third characteristic statistic, N represents the number of samples, represents the fourth characteristic statistic, A cn(n)>at represents a non-weak signal, at represents a threshold level, and E() represents a Gaussian density function.

[0151] Specifically, the first number is actually the A cn (n)>the number of signals with amplitude variation information of at. The second characteristic statistic characterizes the amplitude variation information of the signal within a symbol interval, which can be divided into signal modulation modes in which the normalized central instantaneous amplitude is zero within a symbol interval, such as MPSK signals, and signal modulation modes in which the normalized central instantaneous amplitude is not zero, such as DSB composite signals and AM-FM composite signals. The third characteristic statistic can distinguish signal modulation modes that do not have normalized absolute amplitude information, such as 2ASK signals, and signal modulation modes that have normalized absolute amplitude information, such as high-order MASK signals. The fourth characteristic statistic can distinguish signals with high-density distribution of instantaneous amplitude, such as AM signals, and signals with relatively sparse distribution of instantaneous amplitude, such as MASK signals.

[0152] In the above, when the instantaneous feature set includes the instantaneous phase, the instantaneous feature statistics corresponding to the signal include the fifth feature statistics and the sixth feature statistics, the fifth feature statistics characterizes the standard deviation of the instantaneous phase nonlinear component of the zero-center non-weak signal segment, and the sixth feature statistics characterizes the standard deviation of the absolute value of the instantaneous phase nonlinear component of the zero-center non-weak signal segment. In the above step S103, the instantaneous feature statistics corresponding to the signal are determined according to the instantaneous features, including:

[0153] Perform zero center processing on the instantaneous phase to obtain the nonlinear component of the instantaneous phase;

[0154] The first number and the nonlinear component are respectively input into the fifth feature determination formula and the sixth feature determination formula to obtain the fifth feature statistic and the sixth feature statistic; wherein,

[0155] The fifth characteristic determination formula includes:

[0156]

[0157] The sixth characteristic determination formula includes:

[0158]

[0159] In the above formula, σ dp represents the fifth characteristic statistic, φ NL (n) represents the nonlinear component, C represents the first number, σ ap represents the sixth characteristic statistic, A cn (n)>at indicates a non-weak signal, and at indicates a threshold level.

[0160] Among them, the fifth characteristic statistic characterizes the change of the instantaneous phase of the signal, and can distinguish signals with direct phase information, such as DSB signals, LSB signals, USB signals, 2PSK signals, and signals that do not contain direct phase information, such as AM signals, 2ASK signals, and 4ASK signals. The sixth characteristic statistic can distinguish signals containing absolute phase information, such as 4PSK signals, and signals that do not contain phase information, such as 2ASK signals and 2PSK signals.

[0161] In the above, when the instantaneous feature set includes the instantaneous frequency, the instantaneous feature statistics corresponding to the signal include the seventh feature statistic and the eighth feature statistic, the seventh feature statistic represents the standard deviation of the absolute value of the instantaneous frequency of the zero-centered normalized non-weak signal segment, and the eighth feature statistic represents the compactness of the zero-centered normalized instantaneous frequency. In the above step S103, the instantaneous feature statistics corresponding to the signal are determined according to the instantaneous features, including:

[0162] Processing the instantaneous frequency to obtain a first instantaneous frequency and a second instantaneous frequency, wherein the first instantaneous frequency represents a zero-centered normalized non-weak signal segment instantaneous frequency, and the second instantaneous frequency represents a zero-centered normalized instantaneous frequency;

[0163] Inputting the first number and the first instantaneous frequency into a seventh characteristic determination formula to obtain a seventh characteristic statistic;

[0164] The second instantaneous frequency is input into the eighth characteristic determination formula to obtain the eighth characteristic statistic; wherein,

[0165] The seventh characteristic determination formula includes:

[0166]

[0167] The eighth characteristic determination formula includes:

[0168]

[0169] In the above formula, σ af represents the seventh characteristic statistic, C represents the first number, represents the eighth characteristic statistic, A cn (n)>at indicates a non-weak signal, at indicates the threshold level, E() indicates the Gaussian density function, f N (n) is the first instantaneous frequency, f cn (n) is the second instantaneous frequency.

[0170] The eighth characteristic statistic can distinguish signals with high and dense instantaneous frequency distribution, such as FM signals, and can distinguish signals with relatively sparse instantaneous frequency distribution, such as MFSK signals.

[0171] In the above, the first instantaneous frequency can be determined by:

[0172] Determine the instantaneous frequency mean;

[0173] According to the instantaneous frequency mean and the instantaneous frequency, the instantaneous frequency after zero center processing is determined;

[0174] The first instantaneous frequency is determined according to the instantaneous frequency after zero-center processing.

[0175] Among them, the instantaneous frequency mean can be expressed by the following formula:

[0176]

[0177] The instantaneous frequency after zero centering can be expressed by the following formula:

[0178] f c (n) = f(n) - m f

[0179] The first instantaneous frequency can be expressed by the following formula:

[0180] f N (n) = f c (n)r b

[0181] In the above formula, m f represents the instantaneous frequency mean, N represents the number of samples, f(n) represents the instantaneous frequency, f c (n) represents the instantaneous frequency after zero center processing, f N (n) represents the first instantaneous frequency, r b Represents the normalization coefficient.

[0182] In the above, the second instantaneous frequency can be determined by:

[0183] Determine the instantaneous frequency mean;

[0184] A second instantaneous frequency is determined according to the instantaneous frequency and the instantaneous frequency mean.

[0185] The method for determining the instantaneous frequency mean value may refer to the above description, which will not be described in detail in this embodiment. The second instantaneous frequency may be expressed by the following formula:

[0186]

[0187] In the above formula, f cn (n) is the second instantaneous frequency, m f represents the instantaneous frequency mean, and f(n) represents the instantaneous frequency.

[0188] It should be noted that the process of extracting features from signals can be referred to Figure 2 shown.

[0189] S104: Inputting the instantaneous feature statistic set into the target recognition model so that the target recognition model outputs a target signal modulation mode corresponding to the signal.

[0190] In this embodiment, after obtaining the instantaneous feature statistic set corresponding to the signal, each instantaneous feature statistic in the instantaneous feature statistic set is input into the target recognition model, so that the target recognition model identifies the target signal modulation mode corresponding to the signal according to each instantaneous feature statistic, and when the target recognition model identifies the target signal modulation mode corresponding to the signal, the target signal modulation mode is output.

[0191] Among them, after obtaining the target signal modulation mode, the received signal can be demodulated using the target signal modulation mode to obtain a target demodulation result corresponding to the signal. When the target demodulation result is a demodulation failure, target prompt information corresponding to the signal is generated according to the target demodulation result, and the target prompt information is sent to the target terminal corresponding to the signal. When the target adjustment information corresponding to the target prompt information displayed in the target terminal is received, the target recognition model is updated using the target adjustment information to obtain an updated target recognition model, and the instantaneous feature statistics are input into the target recognition model until the target demodulation result is a demodulation success.

[0192] Specifically, in order to improve the accuracy of the target recognition model for the signal modulation method, when the signal demodulation fails using the recognized target signal modulation method, a corresponding target prompt information is generated, and the target prompt information may be "target signal modulation method recognition error, please readjust the parameters of the target recognition model". After the target prompt information is generated, the target prompt information is pushed to the target terminal, which may be a mobile phone or a computer. After the target terminal receives the target prompt information, the target prompt information is displayed on the target terminal, and the relevant personnel at the target terminal adjust the parameters of the target recognition model based on the target terminal to obtain the target adjustment information corresponding to the target recognition model, thereby updating the target recognition model according to the target adjustment information, and then using the target recognition model to accurately recognize the target signal modulation method corresponding to the signal, so as to achieve accurate demodulation of the signal.

[0193] The present embodiment provides a method for blindly identifying a signal modulation mode. A target recognition model for identifying a signal modulation mode corresponding to a received signal is pre-established. When a signal is received, each instantaneous feature of the signal is extracted. The instantaneous feature statistics corresponding to each instantaneous feature are determined by using the extracted instantaneous features. All the instantaneous feature statistics obtained are input into a pre-trained target recognition model, and then the target recognition model outputs the target signal modulation mode corresponding to the signal. This avoids the drawbacks of blindly identifying a signal modulation mode using a single signal feature, and improves the accuracy and reliability of blindly identifying the signal modulation mode corresponding to the signal.

[0194] refer to Figure 3 , Figure 3 A structural diagram of a blind recognition system for a signal modulation mode provided in an embodiment of the present application. A blind recognition system for a signal modulation mode provided in this embodiment includes a signal sending device 301 and a signal receiving device 302, wherein the signal sending device 301 is used to send a signal to the signal receiving device 302; the signal receiving device 302 is used to establish a target recognition model, and when receiving a signal, execute the blind recognition method of the signal modulation mode as described above, and the target recognition model is used to identify the signal modulation mode corresponding to the received signal.

[0195] refer to Figure 4 , Figure 4 A schematic diagram of the structure of a blind recognition device for a signal modulation mode provided in an embodiment of the present application. A blind recognition device for a signal modulation mode provided in an embodiment of the present application includes: an establishment module 401, an extraction module 402, a determination module 403 and an identification module 404. Among them, the establishment module 401 is used to establish a target recognition model, and the target recognition model is used to identify the signal modulation mode corresponding to the received signal; the extraction module 402 is used to extract each instantaneous feature of the signal when the signal is received to obtain an instantaneous feature set corresponding to the signal; the determination module 403 is used to determine the instantaneous feature statistics corresponding to the signal according to the instantaneous features in the instantaneous feature set for each instantaneous feature in the instantaneous feature set to obtain an instantaneous feature statistic set corresponding to the signal; the identification module 404 is used to input the instantaneous feature statistic set into the target recognition model so that the target recognition model outputs the target signal modulation mode corresponding to the signal.

[0196] In this embodiment, when the instantaneous feature set includes the instantaneous amplitude, the instantaneous feature statistic corresponding to the signal includes a first feature statistic, and the first feature statistic represents the maximum value of the instantaneous amplitude spectral density. The determination module 403 is further used to:

[0197] When the instantaneous feature set includes the instantaneous amplitude, obtaining the number of samples corresponding to the signal;

[0198] Performing zero-center normalization processing on the instantaneous amplitude to obtain the processed instantaneous amplitude;

[0199] The processed instantaneous amplitude and the number of samples are input into a first feature determination formula to obtain the first feature statistic, wherein the first feature determination formula includes:

[0200] γ max =maxDFT(A cn (n)) 2 N

[0201] In the above formula, γ max represents the first characteristic statistic, max represents the maximum value, DFT() represents discrete Fourier transform, A cn (n) represents the instantaneous amplitude after processing, and N represents the number of samples.

[0202] In this embodiment, when the instantaneous feature set includes instantaneous amplitude, the instantaneous feature statistics corresponding to the signal also include a second feature statistic, a third feature statistic and a fourth feature statistic, wherein the second feature statistic represents the standard deviation of the zero-centered normalized non-weak signal segment instantaneous amplitude, the third feature statistic represents the standard deviation of the absolute value of the zero-centered normalized instantaneous amplitude, and the fourth feature statistic represents the compactness of the zero-centered normalized instantaneous amplitude. The determination module 403 is also used to:

[0203] Determine the first number that is not a weak signal from the signals of the sampled number;

[0204] Inputting the first number and the processed instantaneous amplitude into a second feature determination formula to obtain the second feature statistic;

[0205] Inputting the number of samples and the processed instantaneous amplitude into a third feature determination formula to obtain the third feature statistic;

[0206] The processed instantaneous amplitude is input into the fourth feature determination formula to obtain the fourth feature statistic; wherein,

[0207] The second feature determination formula includes:

[0208]

[0209] The third characteristic determination formula includes:

[0210]

[0211] The fourth characteristic determination formula includes:

[0212]

[0213] In the above formula, σ da represents the second characteristic statistic, A cn (n) represents the instantaneous amplitude after processing, C represents the first number, σ aa represents the third characteristic statistic, N represents the number of samples, represents the fourth characteristic statistic, A cn (n)>at represents a non-weak signal, at represents a threshold level, and E() represents a Gaussian density function.

[0214] In this embodiment, when the instantaneous feature set includes an instantaneous phase, the instantaneous feature statistics corresponding to the signal include a fifth feature statistic and a sixth feature statistic, the fifth feature statistic characterizing the standard deviation of the instantaneous phase nonlinear component of the zero-center non-weak signal segment, and the sixth feature statistic characterizing the standard deviation of the absolute value of the instantaneous phase nonlinear component of the zero-center non-weak signal segment. The determination module 403 is also used to:

[0215] When the instantaneous feature set includes the instantaneous phase, performing zero center processing on the instantaneous phase to obtain a nonlinear component of the instantaneous phase;

[0216] The first number and the nonlinear component are respectively input into the fifth feature determination formula and the sixth feature determination formula to obtain the fifth feature statistic and the sixth feature statistic; wherein,

[0217] The fifth characteristic determination formula includes:

[0218]

[0219] The sixth characteristic determination formula includes:

[0220]

[0221] In the above formula, σ dp represents the fifth characteristic statistic, φ NL (n) represents the nonlinear component, C represents the first number, σ ap represents the sixth characteristic statistic, A cn (n)>at indicates a non-weak signal, and at indicates a threshold level.

[0222] In this embodiment, when the instantaneous feature set includes the instantaneous frequency, the instantaneous feature statistics corresponding to the signal include the seventh feature statistic and the eighth feature statistic, the seventh feature statistic characterizes the standard deviation of the absolute value of the instantaneous frequency of the zero-centered normalized non-weak signal segment, and the eighth feature statistic characterizes the compactness of the zero-centered normalized instantaneous frequency. The determination module 403 is also used to:

[0223] When the instantaneous feature set includes the instantaneous frequency, the instantaneous frequency is processed to obtain a first instantaneous frequency and a second instantaneous frequency, wherein the first instantaneous frequency represents a zero-centered normalized non-weak signal segment instantaneous frequency, and the second instantaneous frequency represents a zero-centered normalized instantaneous frequency;

[0224] Inputting the first number and the first instantaneous frequency into a seventh feature determination formula to obtain the seventh feature statistic;

[0225] The second instantaneous frequency is input into the eighth feature determination formula to obtain the eighth feature statistic; wherein,

[0226] The seventh characteristic determination formula includes:

[0227]

[0228] The eighth characteristic determination formula includes:

[0229]

[0230] In the above formula, σ af represents the seventh characteristic statistic, C represents the first number, represents the eighth characteristic statistic, A cn (n)>at indicates a non-weak signal, at indicates the threshold level, E() indicates the Gaussian density function, f N (n) represents the first instantaneous frequency, f cn (n) represents the second instantaneous frequency.

[0231] In this embodiment, the establishing module 401 is further used for:

[0232] Acquire a first preset signal set and a preset signal modulation mode corresponding to each first preset signal in the first preset signal set;

[0233] Performing preset instantaneous feature extraction on each of the first preset signals in the first preset signal set to obtain a preset instantaneous feature set corresponding to each of the preset signals;

[0234] For each of the preset instantaneous features in the preset instantaneous feature set corresponding to each of the first preset signals in the first preset signal set, determining, according to the preset instantaneous features, a preset instantaneous feature statistic corresponding to the preset signal, so as to obtain a preset instantaneous feature statistic set corresponding to each of the first preset signals in the first preset signal set;

[0235] Associating a preset instantaneous feature statistic set corresponding to each of the first preset signals in the first preset signal set with a preset signal modulation mode corresponding to each of the first preset signals in the first preset signal set to obtain a first association relationship;

[0236] The preset machine learning model is trained using the first association relationship to establish the target recognition model.

[0237] In this embodiment, the extraction module 402 is further used for:

[0238] Determine whether the received signal is in a second preset signal set, where a second preset signal in the second preset signal set represents a preset signal of a known signal modulation mode;

[0239] When the received signal is not in the second preset signal set, each instantaneous feature of the signal is extracted to obtain an instantaneous feature set corresponding to the signal.

[0240] In this embodiment, the establishing module 401 is further used for:

[0241] A decision tree model is established, wherein the decision tree model is used to identify a signal modulation mode corresponding to a received signal.

[0242] In this embodiment, the extraction module 402 is further used for:

[0243] When the signal is received, determining whether the received signal is in a third preset signal set, wherein a third preset signal in the third preset signal set represents a preset signal whose signal modulation mode can be identified by using the decision tree model;

[0244] When the received signal is in the third preset signal set, using the decision tree model to identify the target signal modulation mode corresponding to the signal;

[0245] When the received signal is not in the third preset signal set, each instantaneous feature of the signal is extracted to obtain an instantaneous feature set corresponding to the signal.

[0246] The present embodiment provides a blind recognition device for a signal modulation mode, which pre-establishes a target recognition model for identifying the signal modulation mode corresponding to a received signal, so that when a signal is received, feature extraction is performed on each instantaneous feature of the signal, and the instantaneous feature statistics corresponding to each instantaneous feature are determined by using the extracted instantaneous features, so that all the obtained instantaneous feature statistics are input into the pre-trained target recognition model, and then the target recognition model outputs the target signal modulation mode corresponding to the signal, thereby avoiding the drawbacks of blind recognition of the signal modulation mode using a single signal feature, and improving the accuracy and reliability of blind recognition of the signal modulation mode corresponding to the signal.

[0247] like Figure 5 As shown, an embodiment of the present application provides an electronic device, including a processor 501, a communication interface 502, a memory 503 and a communication bus 504, wherein the processor 501, the communication interface 502, and the memory 503 communicate with each other through the communication bus 504.

[0248] Memory 503, used for storing computer programs;

[0249] In one embodiment of the present application, the processor 501 is used to implement the blind identification method of the signal modulation method provided by any one of the aforementioned method embodiments when executing the program stored in the memory 503.

[0250] An embodiment of the present application also provides a computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, the steps of the blind identification method of the signal modulation method provided in any of the aforementioned method embodiments are implemented.

[0251] The device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0252] Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a general hardware platform, and of course, by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the relevant technology can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0253] It should be understood that the terms used herein are only for the purpose of describing specific example embodiments and are not intended to be limiting. Unless the context clearly indicates otherwise, the singular forms "one", "an" and "said" as used herein may also be meant to include plural forms. The terms "include", "comprise", "contain", and "have" are inclusive, and therefore specify the existence of stated features, steps, operations, elements and / or parts, but do not exclude the existence or addition of one or more other features, steps, operations, elements, parts, and / or combinations thereof. The method steps, processes, and operations described herein are not interpreted as necessarily requiring them to be performed in the specific order described or illustrated, unless the execution order is clearly indicated. It should also be understood that additional or alternative steps may be used.

[0254] The foregoing is merely a specific embodiment of the present invention, which enables those skilled in the art to understand or implement the present invention. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but rather to the widest scope consistent with the principles and novel features claimed herein.

Claims

1. A blind identification method of a signal modulation mode, characterized in that: include: Establishing a target recognition model, wherein the target recognition model is used to identify a signal modulation mode corresponding to a received signal; When a signal is received, each instantaneous feature of the signal is extracted to obtain an instantaneous feature set corresponding to the signal; For each of the instantaneous features in the instantaneous feature set, determine an instantaneous feature statistic corresponding to the signal according to the instantaneous feature, so as to obtain an instantaneous feature statistic set corresponding to the signal; The instantaneous feature statistic set is input into the target recognition model so that the target recognition model outputs a target signal modulation mode corresponding to the signal.

2. The method according to claim 1, characterized in that When the instantaneous feature set includes instantaneous amplitude, the instantaneous feature statistic corresponding to the signal includes a first feature statistic, and the first feature statistic represents the maximum value of the instantaneous amplitude spectral density; Determining, according to the instantaneous feature, an instantaneous feature statistic corresponding to the signal includes: When the instantaneous feature set includes the instantaneous amplitude, obtaining the number of samples corresponding to the signal; Performing zero-center normalization processing on the instantaneous amplitude to obtain the processed instantaneous amplitude; The processed instantaneous amplitude and the number of samples are input into a first feature determination formula to obtain the first feature statistic, wherein the first feature determination formula includes: <h2 style=";text-align:left;direction:ltr">γ<h2 style=";text-align:left;direction:ltr"> max <h2 style=";text-align:left;direction:ltr"> =maxDFT(A<h2 style=";text-align:left;direction:ltr"> cn <h2 style=";text-align:left;direction:ltr"> (n))<h2 style=";text-align:left;direction:ltr"> 2 <h2 style=";text-align:left;direction:ltr"> N In the above formula, γ max represents the first characteristic statistic, max represents the maximum value, DFT() represents discrete Fourier transform, A cn (n) represents the instantaneous amplitude after processing, and N represents the number of samples.

3. The method according to claim 2, characterized in that When the instantaneous feature set includes the instantaneous amplitude, the instantaneous feature statistics corresponding to the signal also include a second feature statistic, a third feature statistic and a fourth feature statistic, the second feature statistic characterizes the standard deviation of the zero-centered normalized non-weak signal segment instantaneous amplitude, the third feature statistic characterizes the standard deviation of the absolute value of the zero-centered normalized instantaneous amplitude, and the fourth feature statistic characterizes the compactness of the zero-centered normalized instantaneous amplitude; Determining, according to the instantaneous feature, an instantaneous feature statistic corresponding to the signal includes: Determine the first number that is not a weak signal from the signals of the sampled number; Inputting the first number and the processed instantaneous amplitude into a second feature determination formula to obtain the second feature statistic; Inputting the number of samples and the processed instantaneous amplitude into a third feature determination formula to obtain the third feature statistic; The processed instantaneous amplitude is input into the fourth feature determination formula to obtain the fourth feature statistic; wherein, The second feature determination formula includes: The third characteristic determination formula includes: The fourth characteristic determination formula includes: In the above formula, σ da represents the second characteristic statistic, A cn (n) represents the instantaneous amplitude after processing, C represents the first number, σ aa represents the third characteristic statistic, N represents the number of samples, represents the fourth characteristic statistic, A cn (n)>at represents a non-weak signal, at represents a threshold level, and E() represents a Gaussian density function.

4. The method according to claim 3, characterized in that When the instantaneous feature set includes an instantaneous phase, the instantaneous feature statistics corresponding to the signal include a fifth feature statistic and a sixth feature statistic, the fifth feature statistic characterizing the standard deviation of the instantaneous phase nonlinear component of the zero-center non-weak signal segment, and the sixth feature statistic characterizing the standard deviation of the absolute value of the instantaneous phase nonlinear component of the zero-center non-weak signal segment; Determining, according to the instantaneous feature, an instantaneous feature statistic corresponding to the signal includes: When the instantaneous feature set includes the instantaneous phase, performing zero center processing on the instantaneous phase to obtain a nonlinear component of the instantaneous phase; The first number and the nonlinear component are respectively input into the fifth feature determination formula and the sixth feature determination formula to obtain the fifth feature statistic and the sixth feature statistic; wherein, The fifth characteristic determination formula includes: The sixth characteristic determination formula includes: In the above formula, σ dp represents the fifth characteristic statistic, φ NL (n) represents the nonlinear component, C represents the first number, σ ap represents the sixth characteristic statistic, A cn (n)>at indicates a non-weak signal, and at indicates a threshold level.

5. The method according to claim 3, characterized in that: When the instantaneous feature set includes an instantaneous frequency, the instantaneous feature statistics corresponding to the signal include a seventh feature statistic and an eighth feature statistic, the seventh feature statistic characterizing the standard deviation of the absolute value of the instantaneous frequency of the zero-centered normalized non-weak signal segment, and the eighth feature statistic characterizing the compactness of the zero-centered normalized instantaneous frequency; Determining, according to the instantaneous feature, an instantaneous feature statistic corresponding to the signal includes: When the instantaneous feature set includes the instantaneous frequency, the instantaneous frequency is processed to obtain a first instantaneous frequency and a second instantaneous frequency, wherein the first instantaneous frequency represents a zero-centered normalized non-weak signal segment instantaneous frequency, and the second instantaneous frequency represents a zero-centered normalized instantaneous frequency; Inputting the first number and the first instantaneous frequency into a seventh feature determination formula to obtain the seventh feature statistic; The second instantaneous frequency is input into the eighth feature determination formula to obtain the eighth feature statistic; wherein, The seventh characteristic determination formula includes: The eighth characteristic determination formula includes: In the above formula, σ af represents the seventh characteristic statistic, C represents the first number, represents the eighth characteristic statistic, A cn (n)>at indicates a non-weak signal, at indicates the threshold level, E() indicates the Gaussian density function, f N (n) represents the first instantaneous frequency, f cn (n) represents the second instantaneous frequency.

6. The method according to claim 1, characterized in that The target recognition model is established, comprising: Acquire a first preset signal set and a preset signal modulation mode corresponding to each first preset signal in the first preset signal set; Performing preset instantaneous feature extraction on each of the first preset signals in the first preset signal set to obtain a preset instantaneous feature set corresponding to each of the preset signals; For each of the preset instantaneous features in the preset instantaneous feature set corresponding to each of the first preset signals in the first preset signal set, determining, according to the preset instantaneous features, a preset instantaneous feature statistic corresponding to the preset signal, so as to obtain a preset instantaneous feature statistic set corresponding to each of the first preset signals in the first preset signal set; Associating a preset instantaneous feature statistic set corresponding to each of the first preset signals in the first preset signal set with a preset signal modulation mode corresponding to each of the first preset signals in the first preset signal set to obtain a first association relationship; The preset machine learning model is trained using the first association relationship to establish the target recognition model.

7. The method according to claim 1, characterized in that When receiving a signal, extracting each instantaneous feature of the signal to obtain an instantaneous feature set corresponding to the signal includes: Determine whether the received signal is in a second preset signal set, where a second preset signal in the second preset signal set represents a preset signal of a known signal modulation mode; When the received signal is not in the second preset signal set, each instantaneous feature of the signal is extracted to obtain an instantaneous feature set corresponding to the signal.

8. The method according to claim 1, characterized in that The method further comprises: Establishing a decision tree model, wherein the decision tree model is used to identify a signal modulation mode corresponding to a received signal; Before performing the step of extracting each instantaneous feature of the signal to obtain an instantaneous feature set corresponding to the signal, the method further includes: When the signal is received, determining whether the received signal is in a third preset signal set, wherein a third preset signal in the third preset signal set represents a preset signal whose signal modulation mode can be identified by using the decision tree model; When the received signal is in the third preset signal set, using the decision tree model to identify the target signal modulation mode corresponding to the signal; When the received signal is not in the third preset signal set, the step of extracting each instantaneous feature of the signal to obtain an instantaneous feature set corresponding to the signal is performed.

9. A blind recognition device for a signal modulation mode, characterized in that: include: An establishment module is used to establish a target recognition model, wherein the target recognition model is used to identify a signal modulation mode corresponding to a received signal; An extraction module, used for extracting each instantaneous feature of the signal when receiving the signal, so as to obtain an instantaneous feature set corresponding to the signal; A determination module, configured to determine, for each of the instantaneous features in the instantaneous feature set, an instantaneous feature statistic corresponding to the signal according to the instantaneous feature, so as to obtain an instantaneous feature statistic set corresponding to the signal; The recognition module is used to input the instantaneous feature statistic set into the target recognition model so that the target recognition model outputs the target signal modulation mode corresponding to the signal.

10. A blind recognition system for signal modulation mode, characterized in that: It includes a signal sending device and a signal receiving device; The signal sending device is used to send a signal to the signal receiving device; The signal receiving device is used to establish a target recognition model, and when receiving a signal, executes the blind recognition method of the signal modulation method as described in any one of claims 1 to 8, and the target recognition model is used to identify the signal modulation method corresponding to the received signal.

Images