A communication radiation source identification method, device, equipment and storage medium

Abstract

The application provides a communication radiation source identification method, device and equipment and a storage medium. The method comprises the following steps: determining a background signal in an acquired radiation source signal, acquiring a feature of the radiation source signal and a feature of the background signal, filtering the feature of the background signal from the feature of the radiation source signal to obtain an individual fingerprint feature of the radiation source signal, weakening the influence of the background signal on the radiation source signal, improving the accuracy of the individual fingerprint feature of the radiation source signal, and finally identifying a communication radiation source corresponding to the radiation source signal according to the individual fingerprint feature, so as to ensure the accuracy of the identification result of the communication radiation source.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of radiation source identification, and particularly relates to a communication radiation source identification method and device, equipment and a storage medium. BACKGROUND

[0002] Electromagnetic technology develops rapidly and is widely used, and is one of the high-tech technologies that attract attention. With the rapid development of information technology and social economy, various new electromagnetic radiation source devices have emerged in succession, resulting in an increasingly complex electromagnetic environment.

[0003] Affected by the electromagnetic environment, the radiation source signal will superimpose complex noise in the transmission process. The more complex the electromagnetic environment is, the stronger the superimposed environmental noise is. When the communication radiation source is identified according to the radiation source signal, the identification of the communication radiation source will be directly affected due to the many noise components in the signal, resulting in low accuracy of the communication radiation source identification. SUMMARY

[0004] Based on the defects and deficiencies of the prior art, the present application provides a communication radiation source identification method, device, equipment and storage medium, which can improve the identification accuracy of the communication radiation source.

[0005] The first aspect of the present application provides a communication radiation source identification method, comprising:

[0006] determining the background signal in the acquired radiation source signal, and acquiring the signal characteristics of the radiation source signal and the signal characteristics of the background signal; the background signal includes signal components other than the center signal components of the radiation source signal, and the center signal components include signal components within a set bandwidth range based on the center frequency point of the radiation source signal;

[0007] filtering the signal characteristics of the background signal from the signal characteristics of the radiation source signal to obtain individual fingerprint characteristics of the radiation source signal;

[0008] According to the individual fingerprint characteristics, the communication radiation source corresponding to the radiation source signal is identified.

[0009] Optionally, the determination of the background signal in the acquired radiation source signal comprises:

[0010] extracting the signal of a set bandwidth from the acquired radiation source signal based on the center frequency point of the radiation source signal to obtain a center narrowband signal;

[0011] acquiring the background signal from the signal in the target frequency range of the radiation source signal, and the target frequency range is a frequency range outside the frequency range of the center narrowband signal.

[0012] Optionally, the frequency range of the background signal is adjacent to the frequency range of the center narrowband signal.

[0013] Optionally, the process of obtaining the radiation source signal comprises:

[0014] receiving a high frequency signal transmitted by the communication radiation source;

[0015] converting the high frequency signal into an intermediate frequency signal, and taking the intermediate frequency signal as the radiation source signal.

[0016] Optionally, the process of obtaining the signal feature of the radiation source signal and the signal feature of the background signal comprises:

[0017] calculating the higher-order statistics of the radiation source signal and the higher-order statistics of the background signal based on a Gaussian mixture model;

[0018] taking the higher-order statistics of the radiation source signal as the signal feature of the radiation source signal, and taking the higher-order statistics of the background signal as the signal feature of the background signal.

[0019] Optionally, the process of calculating the higher-order statistics of the radiation source signal and the higher-order statistics of the background signal based on the Gaussian mixture model comprises:

[0020] establishing a signal Gaussian mixture model corresponding to the radiation source signal and a signal Gaussian mixture model corresponding to the background signal based on the Gaussian mixture model and the radiation source signal and the background signal;

[0021] summing the signal Gaussian mixture model parameters corresponding to each signal frame of the radiation source signal according to the signal Gaussian mixture model corresponding to the radiation source signal, to obtain the higher-order statistics of the radiation source signal;

[0022] summing the signal Gaussian mixture model parameters corresponding to each signal frame of the background signal according to the signal Gaussian mixture model corresponding to the background signal, to obtain the higher-order statistics of the background signal.

[0023] Optionally, the process of establishing the signal Gaussian mixture model comprises:

[0024] inputting each signal frame in an input signal into the Gaussian mixture model for iterative training, so that the Gaussian mixture model divides each signal frame in the input signal into at least one signal frame subset, to obtain a signal Gaussian mixture model corresponding to the input signal;

[0025] calculating the class center mean, the Gaussian function weight and the standard deviation vector of each signal frame subset according to the class center mean calculation rule, the Gaussian function weight calculation rule and the standard deviation vector calculation rule of the Gaussian mixture model.

[0026] determine signal Gaussian mixture model parameters corresponding to the input signals according to the class center mean, Gaussian function weight and standard deviation vector of each signal frame subset in the input signals;

[0027] The input signals include radiation source signals or background signals.

[0028] Optionally, the signal features of the radiation source signals include high-order statistics of the radiation source signals, and the signal features of the background signals include high-order statistics of the background signals.

[0029] The signal features of the background signals are filtered out from the signal features of the radiation source signals to obtain individual fingerprint features of the radiation source signals, including:

[0030] The high-order statistics of the radiation source signals and the high-order statistics of the background signals are subjected to variance calculation to obtain at least one group of variance vectors.

[0031] A group of variance vectors is selected from all the variance vectors as the individual fingerprint features of the radiation source signals.

[0032] Optionally, the selection of the group of variance vectors from all the variance vectors as the individual fingerprint features of the radiation source signals includes:

[0033] The smallest group of variance vectors is selected from all the variance vectors as the individual fingerprint features of the radiation source signals.

[0034] Optionally, the communication radiation source corresponding to the radiation source signals is identified according to the individual fingerprint features, including:

[0035] The individual fingerprint features are input into a pre-trained feature classification model to identify the communication radiation source corresponding to the radiation source signals.

[0036] The feature classification model is a model trained by using individual features of radiation sources carrying radiation source identifiers and a neural network.

[0037] The second aspect of the application provides a communication radiation source identification device, including:

[0038] An acquisition module is configured to determine and acquire background signals in acquired radiation source signals, and to acquire signal features of the radiation source signals and signal features of the background signals. The background signals include signal components other than central signal components of the radiation source signals, and the central signal components include signal components within a set bandwidth range based on a central frequency point of the radiation source signals.

[0039] a feature filtering module, configured to filter out signal features of the background signal from signal features of the radiation source signal, to obtain individual fingerprint features of the radiation source signal;

[0040] a radiation source identification module, configured to identify a communication radiation source corresponding to the radiation source signal according to the individual fingerprint features.

[0041] The third aspect of the present application provides a communication radiation source identification device, comprising:

[0042] a memory and a processor;

[0043] The memory is connected with the processor, and is configured to store programs.

[0044] The processor is configured to realize the communication radiation source identification method by running the programs in the memory.

[0045] The fourth aspect of the present application provides a storage medium, characterized in that the storage medium stores a computer program, and the computer program is executed by a processor to realize the communication radiation source identification method.

[0046] The communication radiation source identification method provided by the present application determines the background signal in the obtained radiation source signal, obtains the features of the radiation source signal and the features of the background signal, filters out the features of the background signal from the features of the radiation source signal, obtains the individual fingerprint features of the radiation source signal, reduces the influence of the background signal on the radiation source signal, improves the accuracy of the individual fingerprint features of the radiation source signal, and finally identifies the communication radiation source corresponding to the radiation source signal according to the individual fingerprint features, so as to ensure the accuracy of the communication radiation source identification result. BRIEF DESCRIPTION OF DRAWINGS

[0047] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the drawings needed in the embodiment or prior art description will be briefly introduced. Obviously, the drawings in the following description are only embodiments of the present application, and those skilled in the art can obtain other drawings according to the provided drawings without creative labor.

[0048] Figure 1 is a flowchart of a communication radiation source identification method provided by an embodiment of the present application;

[0049] Figure 2 is a processing flowchart for determining the background signal in the radiation source signal provided by an embodiment of the present application;

[0050] Figure 3 is a background signal extraction schematic diagram provided by an embodiment of the present application;

[0051] Figure 4 is a processing flow schematic diagram provided by an embodiment of the present application for acquiring signal features of a radiation source signal and signal features of a background signal;

[0052] Figure 5 is a processing flow schematic diagram provided by an embodiment of the present application for calculating high-order statistics of a radiation source signal and high-order statistics of a background signal;

[0053] Figure 6 is a processing flow schematic diagram provided by an embodiment of the present application for establishing a signal Gaussian mixture model;

[0054] Figure 7 is a processing flow schematic diagram provided by an embodiment of the present application for determining individual fingerprint features of a radiation source signal;

[0055] Figure 8 is a structural schematic diagram of a communication radiation source identification device provided by an embodiment of the present application;

[0056] Figure 9 is a structural schematic diagram of a communication radiation source identification device provided by an embodiment of the present application. DETAILED DESCRIPTION

[0057] The technical scheme of the embodiment of the present application is applicable to the application scenario of communication radiation source identification, such as identifying a communication radiation source in an electromagnetic environment. By using the technical scheme of the embodiment of the present application, the influence of the electromagnetic environment on the radiation source signal can be weakened, the accuracy of the individual fingerprint features of the radiation source signal can be improved, and the accuracy of the communication radiation source identification result can be ensured.

[0058] As various new electromagnetic radiation source devices have emerged in succession, the electromagnetic environment has become more complex and diverse, and in particular, transient pulse signals and burst signal signals are short-lived, which makes the identification of radiation sources more difficult. Short-wave radiation sources are transmitted through the ionosphere and will superimpose complex electromagnetic environment noise when reflected by the ionosphere. The traditional communication radiation source identification is to increase the length of time to increase the amount of information, thereby weakening the complex noise, but the increase in time affects the efficiency of the communication radiation source identification. In addition, due to the complex and variable signal transmission path in radiation source identification, when the signal receiver point changes or the receiving time period changes, the electromagnetic background environment changes, thereby affecting the target individual of the radiation source, resulting in low accuracy of the individual features in the radiation source signal and affecting the accuracy of the communication radiation source identification result.

[0059] In view of the deficiencies of the prior art described above and the fact that the individual feature accuracy of the radiation source signal is low, which affects the accuracy of the communication radiation source identification result, the present application inventors have conducted research and experiments and propose a communication radiation source identification method, which can weaken the influence of the electromagnetic environment on the radiation source signal, improve the accuracy of the individual fingerprint feature of the radiation source signal, and ensure the accuracy of the communication radiation source identification result.

[0060] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative labor fall within the scope of the present application.

[0061] Exemplary method

[0062] The present application embodiment proposes a communication radiation source identification method, as shown in Figure 1 The method comprises the following steps.

[0063] S101, determine the background signal in the acquired radiation source signal, and acquire the signal features of the radiation source signal and the signal features of the background signal.

[0064] Specifically, the communication radiation source that can emit radio frequency signals includes but is not limited to handheld radios, broadcast stations, intercoms, etc. If the communication radiation source that emits radio frequency signals is to be identified, the present embodiment first needs to receive the radio frequency signals emitted by the communication radiation source to be identified and acquire the radiation source signals corresponding to the radio frequency signals. The specific steps of acquiring the radiation source signals are as follows:

[0065] First, receive the high frequency signals sent by the communication radiation source.

[0066] The radio frequency signals emitted by each communication radiation source are high frequency signals. The present embodiment needs to use a wideband receiver to receive the high frequency signals for communication radiation source identification. The wideband receiver receives the high frequency signals by using a receiving side antenna.

[0067] Second, convert the high frequency signals into intermediate frequency signals, and take the intermediate frequency signals as the radiation source signals.

[0068] In order to enable the amplifier in the wideband receiver to work stably and reduce interference, the present embodiment needs to perform down-conversion processing on the received high frequency signals, thereby converting the high frequency signals into intermediate frequency signals. The intermediate frequency signals are taken as the radiation source signals, and the radiation source signals are identified for the communication radiation source.

[0069] When it is needed to analyze the radio frequency signal emitted by the communication radiation source, for example, to listen to the radio station, it is needed to analyze the radio frequency signal emitted by the radio station into the audio signal so as to be listened by the user, at this time, it is also needed to further convert the converted intermediate frequency signal into the baseband signal (I / Q signal), and then extract the real part information sample from the baseband signal into the audio signal (the sampling frequency is preferably 16K), so as to realize the analysis of the radio frequency signal. The above-mentioned baseband signal can be understood as a complex signal of I+Qj, wherein I: in-phase represents the in-phase, Q represents the quadrature, which is 90 degrees out of phase with I, in the complex function, I is the real part, and J is the imaginary part. The embodiment only identifies the communication radiation source emitting the radio frequency signal, and only needs to use the converted intermediate frequency signal, without further converting into the baseband signal.

[0070] After the radiation source signal is acquired in the embodiment, the background signal is extracted from the radiation source signal. The data at the center frequency point of the radiation source signal contains the most individual characteristics of the radiation source signal, therefore, the frequency domain is divided based on the center frequency point of the radiation source signal, so as to avoid extracting the background signal in the frequency range with more individual characteristics. If the background signal is extracted in the frequency range with more individual characteristics, the important individual characteristics may be filtered out when the signal characteristics of the background signal are filtered out from the signal characteristics of the radiation source signal, which affects the identification of the communication radiation source.

[0071] Therefore, the center signal range is set based on the center frequency point of the radiation source signal in the embodiment, the signal components in the range are the center signal components, and the range of the background signal is extracted based on the above-mentioned bandwidth range division, which is the frequency range outside the center signal range of the radiation source signal. The signal components in the above-mentioned center signal range are the center signal components, and the signal components included in the above-mentioned background signal are the signal components outside the center signal components of the radiation source signal. That is to say, the signal components in the background signal include the noise signal components of the electromagnetic background environment, and also include a part of the signal components corresponding to the individual characteristics, but do not include the signal components with the strongest individual characteristics in the set bandwidth range based on the center frequency point of the radiation source signal.

[0072] After the background signal in the radiation source signal is extracted, the signal feature extraction method is used to obtain the signal feature of the radiation source signal and the signal feature of the background signal. The time domain feature extraction based method such as the time series model method (autoregressive model, autoregressive moving average model) can be used, the frequency domain feature extraction based method such as the fast Fourier transform (FFT) can be used, and the time-frequency domain feature extraction method such as the short-time Fourier transform (STFT), the video distribution (Wigner-Ville distribution, Choi-William distribution), the wavelet transform, the Hilbert-Huang transform, etc. can be used to extract the signal feature of the radiation source signal and the signal feature of the background signal. As a preferred embodiment, the Gaussian mixture model is used to extract the signal feature of the radiation source signal and the signal feature of the background signal.

[0073] S102, the signal feature of the background signal is filtered from the signal feature of the radiation source signal to obtain the individual fingerprint feature of the radiation source signal.

[0074] Specifically, in order to weaken the influence of complex noise in the electromagnetic background environment on the communication radiation source identification, the signal feature of the background signal is filtered from the signal feature of the radiation source signal to obtain the individual fingerprint feature of the radiation source signal.

[0075] The signal feature of the radiation source signal includes the individual feature and the signal feature of the electromagnetic background environment. The signal feature of the background signal includes part of the individual feature and part of the signal feature of the electromagnetic background environment. Based on the above background signal extraction method, the signal feature of the background signal does not include the feature corresponding to the central signal component of the radiation source signal, that is, it does not include the signal component that best reflects the signal source feature. Therefore, filtering the signal feature of the background signal from the signal feature of the radiation source signal can retain the feature corresponding to the central signal component of the radiation source signal, that is, it can make the proportion of the feature component in the signal feature of the radiation source signal that best reflects the signal source feature higher, so that the radiation source signal feature is more conducive to identifying the communication radiation source corresponding to the radiation source signal.

[0076] In this embodiment, the signal characteristics of the background signal are filtered from the signal characteristics of the radiation source signal by using a filtering method to obtain the individual fingerprint characteristics of the radiation source signal. The filtering method that can be used includes Wiener filtering, Kalman filtering, matched filtering, wavelet filtering, etc. In this embodiment, the variance filtering method is preferably used to filter the signal characteristics of the background signal from the signal characteristics of the radiation source signal. The signal characteristics of the radiation source signal can be used as the observation characteristics (equivalent to the random variable in the variance calculation), and the signal characteristics of the background signal can be used as the standard characteristics (equivalent to the mathematical expectation in the variance calculation). The variance is a measure of the deviation between the random variable and its mathematical expectation. Therefore, the deviation of the observation characteristics relative to the standard characteristics is determined by the variance filtering in this embodiment. This deviation is the individual fingerprint characteristics of the radiation source signal, so that the filtering of the signal characteristics is realized by the variance filtering method.

[0077] S103, identifying the communication radiation source corresponding to the radiation source signal according to the individual fingerprint characteristics.

[0078] Specifically, the individual fingerprint characteristics of the radiation source signal are determined, and the individual fingerprint characteristics can be identified and classified by feature recognition to determine the radiation source type corresponding to the individual fingerprint characteristics, so as to identify the communication radiation source corresponding to the radiation source signal.

[0079] Further, the specific execution steps of this step are as follows:

[0080] The individual fingerprint characteristics are input into the pre-trained feature classification model to identify the communication radiation source corresponding to the radiation source signal.

[0081] The feature classification model can train the neural network by using the radiation source individual characteristics carrying the radiation source identifier. When identifying the radiation source individual characteristics, the neural network can calculate the similarity scores of the radiation source individual characteristics and each category contained in the neural network classification module. The total score is 1 point. If all the similarity scores are lower than the threshold value (preferably 0.1 points), it means that the neural network classification module does not contain the category corresponding to the radiation source individual characteristics. At this time, other categories are created, the category is put into the other categories, and it is determined that the category matches the radiation source identifier carried by the radiation source individual characteristics. In this way, the registration function of the radiation source can be realized, and the category corresponding to the radiation source is created in the neural network classification module.

[0082] If all the calculated similarity scores have scores higher than the threshold value, the category corresponding to the highest score in all the scores higher than the threshold value is output as the communication radiation source corresponding to the radiation source individual characteristics. Then, the neural network is adjusted according to the verification of the output of the neural network and the radiation source identifier carried by the radiation source individual characteristics, so as to realize the model training, and the trained neural network is used as the feature classification model.

[0083] The process of obtaining the individual feature of the radiation source comprises: obtaining a sample radiation source signal, extracting a sample background signal from the sample radiation source signal, obtaining a signal feature of the sample radiation source signal and a signal feature of the sample background signal, and filtering the signal feature of the sample background signal from the signal feature of the sample radiation source signal to obtain the individual feature of the sample radiation source.

[0084] In addition, in this embodiment, the neural network can be a residual network (ResNet), an Inception network, a convolutional neural network (CNN), a recurrent neural network (RNN), a deep neural network (DNN), a DenseNet network, a VGG network, a GoogleNet network, a long short-term memory artificial neural network (LSTM), an AlexNet network, etc., and preferably a residual network. The residual network is easy to optimize and can improve the accuracy by increasing the depth. The residual block in the residual network uses a skip connection to alleviate the gradient vanishing problem caused by increasing the depth of the deep neural network.

[0085] As can be seen from the above description, the communication radiation source identification method provided in the embodiments of the present application determines the background signal in the obtained radiation source signal, obtains the feature of the radiation source signal and the feature of the background signal, filters the feature of the background signal from the feature of the radiation source signal to obtain the individual fingerprint feature of the radiation source signal, weakens the influence of the background signal on the radiation source signal, improves the accuracy of the individual fingerprint feature of the radiation source signal, and finally identifies the communication radiation source corresponding to the radiation source signal according to the individual fingerprint feature, which can ensure the accuracy of the communication radiation source identification result.

[0086] As an optional implementation, referring to FIG. 1, Figure 2 As shown in FIG. 1, another embodiment of the present application discloses that, in step S101, the background signal in the obtained radiation source signal is determined, comprising:

[0087] S201, taking the center frequency point of the radiation source signal as a reference, extracting a signal with a set bandwidth from the obtained radiation source signal to obtain a center narrowband signal.

[0088] Specifically, as shown in FIG. 1, in order to extract the background signal from the radiation source signal, the center frequency point of the radiation source signal is taken as a reference, the center frequency point is Figure 3 the frequency point where the dashed line a in FIG. 1 is located, and a narrowband filtering method is used to extract a signal with a set bandwidth centered on the above center frequency point from the radiation source signal, and the signal is taken as a center narrowband signal. The signal component contained in the center narrowband signal is a center signal component. Figure 3 Figure 3 ​A represents the radiation source signal, and B represents the center narrowband signal, which can be obtained by filtering the radiation source signal according to a set bandwidth by using a narrow filter. The narrow filter is defined according to the amplitude-frequency characteristic of the frequency characteristic of the filter. If a pole filter is set at the signal frequency to be extracted by the digital filter, a very narrow passband and a relatively steep transition band can be obtained, and therefore the filter is called a narrow-band filter. As shown in Figure 3 Preferably, the radiation source signal A is a 6KHz signal, and the center narrowband signal B is a 1KHz signal.

[0089] S202, obtaining a background signal from a frequency range outside a frequency range in which the center narrowband signal of the radiation source signal is located.

[0090] Specifically, the frequency range in which the center narrowband signal of the radiation source signal is located contains more individual characteristics of the radiation source. If the background signal is extracted in the frequency range, the signal characteristics of the background signal contain more individual characteristics of the radiation source. After the signal characteristics of the background signal in the signal characteristics of the radiation source signal are filtered out, the individual characteristics will be greatly lost, and therefore the accuracy of the radiation source identification will be affected. Therefore, in order to avoid great loss of individual characteristics, the target frequency range for obtaining the background signal should be a frequency range outside the frequency range in which the center narrowband signal of the radiation source signal is located.

[0091] However, in order to make the signal characteristics of the obtained background signal contain more signal characteristics of the electromagnetic background environment, the frequency range for obtaining the background signal should be as close as possible to the center frequency point of the radiation source signal, so as to ensure that as many signal characteristics of the radiation source signal as possible are filtered out of the noise characteristics generated by the electromagnetic background environment. Therefore, the frequency range in which the background signal is located is adjacent to the frequency range in which the center narrowband signal is located, so as to ensure that the frequency range in which the background signal is located is the frequency domain closest to the center frequency point of the radiation source signal except the frequency range in which the center narrowband signal is located.

[0092] Exemplarily, as shown in Figure 3The frequency range of the background signal C is adjacent to the frequency range of the center narrowband signal B. If the bandwidth of the background signal to be acquired is 1KHz and the bandwidth of the center narrowband signal is also 1KHz, then the center frequency point b of the background signal is 1KHz left or right of the center frequency point a of the radiation source signal, and then the signal with a bandwidth of 1KHz is extracted from the radiation source signal with the center frequency point b as the reference to obtain the background signal. If the bandwidth of the background signal to be acquired is 2KHz and the bandwidth of the center narrowband signal is 1KHz, then the center frequency point b of the background signal is 1.5KHz left or right of the center frequency point a of the radiation source signal, and then the signal with a bandwidth of 2KHz is extracted from the radiation source signal with the center frequency point b as the reference to obtain the background signal. If the bandwidth of the background signal to be acquired is 1KHz and the bandwidth of the center narrowband signal is 3KHz, then the center frequency point b of the background signal is 2KHz left or right of the center frequency point a of the radiation source signal, and then the signal with a bandwidth of 1KHz is extracted from the radiation source signal with the center frequency point b as the reference to obtain the background signal.

[0093] As an optional implementation, referring to FIG. 1, Figure 4 As shown in FIG. 1, another embodiment of the present application discloses that, in step 101, the signal features of the radiation source signal and the signal features of the background signal are acquired by:

[0094] S401, based on the Gaussian mixture model, calculating the high-order statistics of the radiation source signal and the high-order statistics of the background signal.

[0095] Specifically, the signal features of the radiation source signal and the signal features of the background signal can be acquired by using the Gaussian mixture model. The Gaussian mixture model can directly fit the statistical distribution of the signal. In this embodiment, the signal Gaussian mixture model of the radiation source signal and the signal Gaussian mixture model of the background signal are obtained by training the Gaussian mixture model. The high-order statistics are calculated by using each signal frame of the radiation source signal and the signal Gaussian mixture model corresponding to the radiation source signal, and a vector group corresponding to the radiation source signal is obtained as the high-order statistics of the radiation source signal. The high-order statistics are calculated by using each signal frame of the background signal and the signal Gaussian mixture model corresponding to the background signal, and a vector group corresponding to the background signal is obtained as the high-order statistics of the background signal.

[0096] The Gaussian mixture model can divide the input signal frame, divide all signal frames into at least one signal frame subset, that is, divide the signal frames close to each other into the same subset. Therefore, the established signal Gaussian mixture model includes at least one signal frame subset, each signal frame subset is a single Gaussian model, and each single Gaussian model corresponds to a calculated model parameter. The model parameter of the signal mixture Gaussian model is calculated according to each single Gaussian model and the model parameter corresponding to each single Gaussian model, so that the corresponding higher-order statistics can be obtained.

[0097] S402, the higher-order statistics of the radiation source signal is taken as the signal feature of the radiation source signal, and the higher-order statistics of the background signal is taken as the signal feature of the background signal.

[0098] Specifically, the higher-order statistics calculated by using different signals to establish the Gaussian mixture model are different, so that the higher-order statistics can be taken as the signal feature to filter and identify the signal. In this embodiment, the higher-order statistics of the radiation source signal is taken as the signal feature of the radiation source signal, and the higher-order statistics of the background signal is taken as the signal feature of the background signal.

[0099] As an optional implementation, referring to FIG. 4, another embodiment of the present application discloses that the step S401 includes: Figure 5

[0100] S501, based on the Gaussian mixture model and the radiation source signal and the background signal, a signal Gaussian mixture model corresponding to the radiation source signal and a signal Gaussian mixture model corresponding to the background signal are established.

[0101] Specifically, the radiation source signal can be input into the Gaussian mixture model, so that the Gaussian mixture model divides all signal frames contained in the radiation source signal into at least one signal frame subset. The division of the signal frames into at least one signal frame subset can be realized by clustering the input signal frames. Each signal frame subset can be a single Gaussian model, and all single Gaussian models constructed by all signal frames in the radiation source signal form the signal Gaussian mixture model corresponding to the radiation source signal. The signal Gaussian mixture model corresponding to the background signal is established in the same way as the signal Gaussian mixture model corresponding to the radiation source signal. The background signal is input into the Gaussian mixture model, and the signal Gaussian mixture model corresponding to the background signal can be established.

[0102] Further, the signal Gaussian mixture model has corresponding signal Gaussian mixture model parameters, wherein the signal Gaussian mixture model parameters include Gaussian function weights, class center mean values and standard deviation vectors. The signal Gaussian mixture model parameters can be represented as: ​

[0103]

[0104] where λ represents the signal Gaussian mixture model parameters, j represents the jth single Gaussian model in the signal Gaussian mixture model, i.e., the jth signal frame subset, M represents the number of single Gaussian models in the signal Gaussian mixture model, i.e., the number of signal frame subsets, and m represents the number of iterations when the signal Gaussian mixture model is completed, represents the Gaussian function weight of the jth single Gaussian model, represents the class center mean of the jth single Gaussian model, represents the standard deviation vector of the jth single Gaussian model. Each single Gaussian model in the signal Gaussian mixture model has a set of model parameters, and the combination of all model parameters is the signal Gaussian mixture model parameter.

[0105] S502, according to the signal Gaussian mixture model corresponding to the radiation source signal, summing the signal Gaussian mixture model parameters corresponding to each signal frame of the radiation source signal to obtain the high-order statistics of the radiation source signal.

[0106] Specifically, after the signal Gaussian mixture model is completed, the signal high-order statistics need to be extracted on the basis of the signal Gaussian mixture model. In this embodiment, first, the signal Gaussian mixture model parameters corresponding to each signal frame need to be determined, and then the signal Gaussian mixture model parameters corresponding to each signal frame of the radiation source signal are summed. The signal Gaussian mixture model parameters include the Gaussian function weight, the class center mean and the standard deviation vector, so the summing operation is to sum the three parameters respectively.

[0107] The summing formula of the Gaussian function weight is as follows:

[0108]

[0109] where n i represents the sum of the Gaussian function weights, T represents the number of all signal frames in the signal Gaussian mixture model, x i represents the ith signal frame, and l represents the single Gaussian model in which the signal frame x i is located, p(l / x i , λ) represents the probability of the Gaussian function weight in the single Gaussian model in which the signal frame x i is located in all Gaussian function weights of the single Gaussian model, and also represents the Gaussian function weight corresponding to the signal frame x i .

[0110] The summing formula of the class center mean is as follows:

[0111]

[0112] where E i represents the sum of the class center means, T represents the number of all signal frames in the signal Gaussian mixture model, x i represents the ith signal frame, and l represents the single Gaussian model in which the signal frame x i is located, p(l / x i , λ) represents the probability of the Gaussian function weight in the single Gaussian model in which the signal frame x i is located in all Gaussian function weights of the single Gaussian model, and also represents the Gaussian function weight corresponding to the signal frame x i .i (x) represents the sum of the class center means, p(l / x i , λ) represents the probability of the class center mean in the single Gaussian model in which the signal frame x i is located in all class center means of the single Gaussian models, x i p(l / x i , λ) represents the corresponding class center mean of the signal frame x i .

[0113] The sum formula of the standard deviation vector is as follows:

[0114]

[0115] wherein, E i (x 2 ) represents the sum of the standard deviation vectors, p(l / x i , λ) represents the probability of the standard deviation vector in the single Gaussian model in which the signal frame x i is located in all standard deviation vectors of the single Gaussian models, x i 2 p(l / x i , λ) represents the corresponding standard deviation vector of the signal frame x i .

[0116] The sum of the Gaussian function weights, the sum of the class center means and the sum of the standard deviation vectors calculated above are combined into a vector group Z, and the vector group Z is taken as the higher-order statistics of the radiation source signal, that is, Z= [n i , E i (x), E i (x 2 )].

[0117] S503, according to the signal Gaussian mixture model corresponding to the background signal, summing the signal Gaussian mixture model parameters corresponding to each signal frame of the background signal to obtain the higher-order statistics of the background signal.

[0118] Specifically, in the embodiment, first, the signal Gaussian mixture model parameters corresponding to each signal frame in the background signal need to be determined, and then the signal Gaussian mixture model parameters corresponding to each signal frame in the background signal are summed to obtain the higher-order statistics of the background signal. The calculation method of the higher-order statistics of the background signal is the same as that of the higher-order statistics of the radiation source signal, and this step will not be described in detail.

[0119] In the embodiment, the execution order of S502 and step S503 is not limited, that is, step S502 can be executed first, and then step S503 can be executed, or step S503 can be executed first, and then step S502 can be executed, or the two steps can be executed simultaneously.

[0120] As an optional implementation, refer to Figure 6 The application further discloses a method for establishing a signal Gaussian mixture model, as shown in the accompanying drawings.

[0121] S601, input each signal frame in the input signal into the Gaussian mixture model for iterative training, so that the Gaussian mixture model divides each signal frame in the input signal into at least one signal frame subset, and obtains a signal Gaussian mixture model corresponding to the input signal.

[0122] Specifically, the input signal can be represented as X=x t , t=1, 2...T, wherein x t represents a signal frame contained in the input signal X, and T represents the number of signal frames in the input signal. Each signal frame in the input signal is input into the Gaussian mixture model for iterative training. After the Gaussian mixture model receives the signal frame, the current iteration number is recorded. Each time a signal frame is input, the current iteration number is increased by one. Then, the signal frame is divided into subsets by using the nearest neighbor criterion formula, and the total iteration distortion value after this iteration is calculated according to the distortion calculation formula, wherein the iteration distortion value of each signal frame subset is the distance from all signal frames in the signal frame subset to the center of the signal frame subset, and the total iteration distortion value is the sum of the iteration distortion values of all signal frame subsets. Then, the change value of the current iteration distortion is calculated according to the distortion improvement calculation formula, the total iteration distortion value after this iteration and the total iteration distortion value after the last iteration. When the current iteration number reaches the maximum iteration number or the change value of the current iteration distortion reaches the pre-set improvement threshold, the iteration is stopped, and each signal frame subset after the division is a single Gaussian model. All single Gaussian models constitute a signal Gaussian mixture model.

[0123] The signal Gaussian mixture model of the radiation source signal is established by taking the radiation source signal as the input signal, and the signal Gaussian mixture model of the background signal is established by taking the background signal as the input signal.

[0124] The nearest neighbor criterion formula is: wherein m represents the current iteration number, represents the class center mean value of the lth signal frame subset at the m-1th iteration, represents the class center mean value of the ith signal frame subset at the m-1th iteration, represents the distance between the input signal frame x and the center of the lth signal frame subset at the mth iteration, represents the distance between the input signal frame x at the mth iteration and the center of the ith signal frame subset. In the formula, if the distance between the input signal frame x at the mth iteration and the center of the lth signal frame subset is greater than the distance between the input signal frame x at the mth iteration and the center of the ith signal frame subset, the input signal frame x at the mth iteration needs to be divided into the lth signal frame subset.

[0125] The distortion calculation formula is:

[0126]

[0127] wherein D (m) represents the total iteration distortion value after the mth iteration; represents the lth signal frame subset at the mth iteration; represents the sum of the distances between all signal frames in the lth signal frame subset and the center of the signal frame subset after the mth iteration.

[0128] The distortion improvement calculation formula is:

[0129]

[0130] wherein σ (m) represents the change value of the iteration distortion after the mth iteration; D (m-1) represents the total iteration distortion value after the mth iteration; D represents the total iteration distortion value after the mth iteration.

[0131] S602, according to the class center mean calculation rule, the Gaussian function weight calculation rule and the standard deviation vector calculation rule of the Gaussian mixture model, the class center mean, the Gaussian function weight and the standard deviation vector of each signal frame subset are calculated.

[0132] Specifically, after the signal Gaussian mixture model corresponding to the input signal is established, the model parameters of the signal Gaussian mixture model need to be calculated by using the class center mean calculation rule, the Gaussian function weight calculation rule and the standard deviation vector calculation rule, wherein the model parameters include the class center mean, the Gaussian function weight and the standard deviation vector of the signal frame subset.

[0133] The class center mean calculation rule is:

[0134]

[0135] wherein, represents the class center mean of the jth signal frame subset, represents the jth signal frame subset, N j represents the number of signal frames contained in the jth signal frame subset, x represents the sum of the data of all signal frames in the jth signal frame subset.

[0136] The rules for calculating the weights of the Gaussian function are as follows:

[0137]

[0138] in, Let represent the Gaussian function weight of the j-th signal frame subset, and T represent the total number of signal frames contained in the input signal.

[0139] The standard deviation vector is calculated according to the following rules:

[0140]

[0141] in, Let represent the standard deviation vector of the j-th signal frame subset.

[0142] In all the above calculation rules, m indicates that the Gaussian mixture model corresponding to the input signal is completed when the Gaussian mixture model is iteratively trained to the mth time.

[0143] S603. Determine the parameters of the Gaussian mixture model corresponding to the input signal based on the mean of the class center, the weight of the Gaussian function, and the standard deviation vector of each signal frame subset in the input signal.

[0144] After calculating the class center mean, Gaussian function weights, and standard deviation vector of each signal frame subset in the input signal using the above calculation rules, the parameters of the signal Gaussian mixture model corresponding to the input signal are determined. The signal Gaussian mixture model parameters can be expressed as:

[0145]

[0146] Where λ represents the parameters of the Gaussian mixture model of the signal, j represents the j-th single Gaussian model in the Gaussian mixture model of the signal, i.e., the j-th subset of signal frames, M represents the number of single Gaussian models in the Gaussian mixture model of the signal, i.e., the number of subsets of signal frames, and m represents the number of iterations when establishing the Gaussian mixture model of the signal. This represents the Gaussian function weights of the j-th single Gaussian model. This represents the class center mean of the j-th single Gaussian model. Let represent the standard deviation vector of the j-th single Gaussian model. Each single Gaussian model in a signal mixture Gaussian model has a set of model parameters, and the combination of all model parameters constitutes the parameters of the signal Gaussian mixture model.

[0147] As an optional implementation method, see [link to implementation details]. Figure 7 As shown, another embodiment of this application discloses that, in step S102 above, filtering out the signal features of the background signal from the signal features of the radiation source signal to obtain the individual fingerprint features of the radiation source signal includes:

[0148] S701, variance calculation is performed on the high-order statistics of the radiation source signal and the high-order statistics of the background signal to obtain at least one group of variance vectors.

[0149] Specifically, the embodiment can realize the statistics filtering through the variance calculation between the high-order statistics, take the high-order statistics of the radiation source signal as an observation vector, take the high-order statistics of the background signal as a standard vector, calculate the variance between the high-order statistics of the radiation source signal and the high-order statistics of the background signal, and obtain a plurality of vectors since the high-order statistics of the radiation source signal and the high-order statistics of the background signal are both vector groups, i.e., matrix groups, and the variance between the matrix groups is a plurality of matrices. The embodiment takes the variance between the high-order statistics of the radiation source signal and the high-order statistics of the background signal as a variance vector, and performs the variance calculation on the high-order statistics of the radiation source signal and the high-order statistics of the background signal, so that at least one group of variance vectors can be obtained.

[0150] Through the variance calculation between the high-order statistics of the radiation source signal and the high-order statistics of the background signal, the influence of the noise signal in the electromagnetic background environment on the radiation source signal can be weakened, the individual characteristics in the radiation source signal are highlighted, and the identification accuracy of the communication radiation source is improved.

[0151] S702, a group of variance vectors is selected from all the variance vectors as the individual fingerprint characteristics of the radiation source signal.

[0152] The embodiment needs to select a group of variance vectors from the variance vectors between the high-order statistics of the radiation source signal and the high-order statistics of the background signal as the individual fingerprint characteristics of the radiation source signal. In the embodiment, the size of the variance vector between the high-order statistics of the radiation source signal and the high-order statistics of the background signal reflects the closeness of the signal characteristics of the radiation source signal and the signal characteristics of the background signal calculated from the group of variance vectors. The smaller the variance vector is, the closer the signal characteristics of the radiation source signal and the signal characteristics of the background signal calculated from the variance vector are, and the more the signal characteristics of the noise signal in the electromagnetic background environment contained in the signal characteristics of the background signal at this time. The characteristics corresponding to the variance vector can better reflect the individual characteristics of the radiation source. Therefore, the embodiment preferably selects the smallest group of variance vectors from all the variance vectors as the individual fingerprint characteristics of the radiation source signal.

[0153] Exemplary apparatus

[0154] Correspondingly, the embodiment of the application further provides a communication radiation source identification device, as shown in Figure 8 The device comprises:

[0155] The acquisition module 100 is configured to determine a background signal in the acquired radiation source signal, and acquire a signal feature of the radiation source signal and a signal feature of the background signal; the background signal includes a signal component other than a central signal component of the radiation source signal, and the central signal component includes a signal component within a set bandwidth range based on a center frequency point of the radiation source signal.

[0156] The feature filtering module 110 is configured to filter the signal feature of the background signal from the signal feature of the radiation source signal to obtain an individual fingerprint feature of the radiation source signal.

[0157] The radiation source identification module 120 is configured to identify a communication radiation source corresponding to the radiation source signal according to the individual fingerprint feature.

[0158] The communication radiation source identification device provided by the embodiment of the present application determines the background signal in the acquired radiation source signal by using the acquisition module 100, and acquires the feature of the radiation source signal and the feature of the background signal, then filters the feature of the background signal from the feature of the radiation source signal by using the feature filtering module 110 to obtain the individual fingerprint feature of the radiation source signal, which reduces the influence of the background signal on the radiation source signal, improves the accuracy of the individual fingerprint feature of the radiation source signal, and finally identifies the communication radiation source corresponding to the radiation source signal according to the individual fingerprint feature by using the radiation source identification module 120, which can ensure the accuracy of the identification result of the communication radiation source.

[0159] As an optional implementation, the other embodiment of the present application further discloses that the acquisition module 100 includes a first signal acquisition unit and a second signal acquisition unit.

[0160] The first signal acquisition unit is configured to extract a signal of a set bandwidth from the acquired radiation source signal based on a center frequency point of the radiation source signal to obtain a central narrowband signal.

[0161] The second signal acquisition unit is configured to acquire a signal from a target frequency range of the radiation source signal to obtain the background signal, and the target frequency range is a frequency range outside a frequency range in which the central narrowband signal is located.

[0162] As an optional implementation, the other embodiment of the present application further discloses that the frequency range in which the background signal is located is adjacent to the frequency range in which the central narrowband signal is located.

[0163] As an optional implementation, the other embodiment of the present application further discloses that the acquisition process of the radiation source signal includes:

[0164] receiving a high-frequency signal sent by the communication radiation source;

[0165] converting the high-frequency signal into an intermediate-frequency signal, and taking the intermediate-frequency signal as the radiation source signal.

[0166] As an optional implementation, the other embodiment of the present application further discloses that the acquisition module 100 further comprises a calculation unit and a determination unit.

[0167] The calculation unit is configured to calculate the high-order statistics of the radiation source signal and the high-order statistics of the background signal based on the Gaussian mixture model.

[0168] The determination unit is configured to take the high-order statistics of the radiation source signal as the signal feature of the radiation source signal, and take the high-order statistics of the background signal as the signal feature of the background signal.

[0169] As an optional implementation, the other embodiment of the present application further discloses that the calculation unit is specifically configured to:

[0170] establish a signal Gaussian mixture model corresponding to the radiation source signal and a signal Gaussian mixture model corresponding to the background signal based on the Gaussian mixture model and the radiation source signal and the background signal;

[0171] sum the signal Gaussian mixture model parameters corresponding to each signal frame of the radiation source signal according to the signal Gaussian mixture model corresponding to the radiation source signal, to obtain the high-order statistics of the radiation source signal;

[0172] sum the signal Gaussian mixture model parameters corresponding to each signal frame of the background signal according to the signal Gaussian mixture model corresponding to the background signal, to obtain the high-order statistics of the background signal.

[0173] As an optional implementation, the other embodiment of the present application further discloses that the establishment process of the signal Gaussian mixture model comprises:

[0174] input each signal frame in the input signal into the Gaussian mixture model for iterative training, so that the Gaussian mixture model divides each signal frame in the input signal into at least one signal frame subset, to obtain the signal Gaussian mixture model corresponding to the input signal;

[0175] calculate the class center mean, the Gaussian function weight and the standard deviation vector of each signal frame subset according to the class center mean calculation rule, the Gaussian function weight calculation rule and the standard deviation vector calculation rule of the Gaussian mixture model;

[0176] determine the signal Gaussian mixture model parameters corresponding to the input signal according to the class center mean, the Gaussian function weight and the standard deviation vector of each signal frame subset in the input signal;

[0177] The input signal comprises the radiation source signal or the background signal.

[0178] As an optional implementation, another embodiment of the present application further discloses that the signal feature of the radiation source signal comprises a high-order statistic of the radiation source signal, and the signal feature of the background signal comprises a high-order statistic of the background signal.

[0179] The feature filtering module 110 is specifically used for:

[0180] The high-order statistics of the radiation source signal and the high-order statistics of the background signal are subjected to variance calculation to obtain at least one group of variance vectors.

[0181] A group of variance vectors is selected from all the variance vectors as individual fingerprint features of the radiation source signal.

[0182] As an optional implementation, another embodiment of the present application further discloses that the group of variance vectors is selected from all the variance vectors as the individual fingerprint features of the radiation source signal, comprising:

[0183] The smallest group of variance vectors is selected from all the variance vectors as the individual fingerprint features of the radiation source signal.

[0184] As an optional implementation, another embodiment of the present application further discloses that the radiation source identification module 120 is specifically used for:

[0185] The individual fingerprint features are input into a pre-trained feature classification model to identify a communication radiation source corresponding to the radiation source signal.

[0186] The feature classification model is a model obtained by training a neural network using individual features of a radiation source carrying a radiation source identifier.

[0187] The communication radiation source identification device provided by the embodiment belongs to the same application concept as the communication radiation source identification method provided by the above-mentioned embodiments of the present application, can execute the communication radiation source identification method provided by any of the above-mentioned embodiments of the present application, and has the corresponding function modules and beneficial effects of executing the communication radiation source identification method. Technical details not described in detail in the embodiment can be referred to the specific processing content of the communication radiation source identification method provided by the above-mentioned embodiments of the present application, which will not be described here.

[0188] Exemplary electronic device

[0189] Another embodiment of the present application further provides a communication radiation source identification device, which is shown in Figure 9 The device comprises:

[0190] a memory 200 and a processor 210;

[0191] The memory 200 is connected with the processor 210, and is used for storing a program.

[0192] The processor 210 is configured to implement the communication radiation source identification method disclosed in any of the above embodiments by running the program stored in the memory 200.

[0193] Specifically, the communication radiation source identification device can further include a bus, a communication interface 220, an input device 230, and an output device 240.

[0194] The processor 210, the memory 200, the communication interface 220, the input device 230, and the output device 240 are connected to each other through the bus.

[0195] The bus can include a path for transmitting information between the various components of the computer system.

[0196] The processor 210 can be a general-purpose processor, such as a central processing unit (CPU), a microprocessor, or the like, an application-specific integrated circuit (ASIC), or one or more integrated circuits for controlling the execution of programs of the present application. It can also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or other programmable logic device, a discrete gate or transistor logic device, a discrete hardware component.

[0197] The processor 210 can include a main processor and can further include a baseband chip, a modem, and the like.

[0198] The memory 200 stores programs for implementing the technical solutions of the present application, and can also store operating systems and other key services. Specifically, the program can include program code, and the program code includes computer operation instructions. More specifically, the memory 200 can include read-only memory (ROM), other types of static storage devices that can store static information and instructions, random access memory (RAM), other types of dynamic storage devices that can store information and instructions, disk storage, flash, and the like.

[0199] The input device 230 can include a device that receives data and information input by a user, such as a keyboard, a mouse, a camera, a scanner, a light pen, a voice input device, a touch screen, a pedometer, or a gravity sensor, and the like.

[0200] The output device 240 can include a device that allows information to be output to a user, such as a display screen, a printer, a speaker, and the like.

[0201] The communication interface 220 can include any transceiver-type device for communicating with other devices or communication networks, such as an Ethernet network, a radio access network (RAN), a wireless local area network (WLAN), etc.

[0202] The processor 210 executes the programs stored in the memory 200, and invokes other devices, which can be used to implement each step of any of the communication radiation source identification methods provided by the embodiments described above.

[0203] Exemplary computer program product and storage medium

[0204] In addition to the above-mentioned methods and devices, the embodiments of the present application can also be computer program products, which include computer program instructions, which, when executed by a processor, cause the processor to perform the steps of the communication radiation source identification methods described in the above "Exemplary Methods" section of the specification.

[0205] The computer program product can be written in any combination of one or more programming languages, including an object-oriented programming language such as Java, C++, etc., and conventional procedural programming languages, such as the "C" programming language, or similar programming languages. The program code can execute entirely on the user's computing device, partly on the user's device, as a stand-alone software package, partly on the user's computing device and partly on a remote computing device or entirely on the remote computing device or server.

[0206] In addition, the embodiments of the present application can also be storage media, which store computer programs, and the computer programs are executed by a processor to perform the steps of the communication radiation source identification methods described in the above "Exemplary Methods" section of the specification.

[0207] For each of the above-mentioned method embodiments, in order to simply describe, they are all expressed as a combination of a series of actions, but those skilled in the art should know that the present application is not limited to the order of the actions described, because according to the present application, certain steps can be performed in other orders or simultaneously. Secondly, those skilled in the art should know that the embodiments described in the specification are all preferred embodiments, and the actions and modules involved are not necessarily required by the present application.

[0208] It should be noted that each of the embodiments in the specification is described in a progressive manner, and each embodiment focuses on the differences from other embodiments, and the same and similar parts between the embodiments can be referred to each other. For device embodiments, since they are basically similar to method embodiments, they are described more simply, and the relevant parts are referred to the part of the description of the method embodiments.

[0209] The steps in the methods of the embodiments of the present application can be adjusted in sequence, combined, and deleted according to actual needs.

[0210] The modules and sub-modules in the devices and terminals in the embodiments of the present application can be combined, divided, and deleted according to actual needs.

[0211] In several embodiments provided in the present application, it should be understood that the disclosed terminals, devices, and methods can be implemented in other ways. For example, the terminal embodiments described above are only schematic. For example, the division of the modules or sub-modules is only a logical function division. In actual implementation, another division mode can be used. For example, a plurality of sub-modules or modules can be combined or integrated into another module, or some features can be ignored or not executed. In addition, the coupling or direct coupling or communication connection between the displayed or discussed mutual elements can be indirect coupling or communication connection through some interfaces, devices, or modules, and can be electrical, mechanical, or other forms.

[0212] The modules or sub-modules described as separate components can or can not be physically separated, and the components of the modules or sub-modules can or can not be physical modules or sub-modules, that is, can be located in one place or can be distributed to a plurality of network modules or sub-modules. Some or all of the modules or sub-modules can be selected according to actual needs to achieve the purpose of the embodiments.

[0213] In addition, each functional module or sub-module in the embodiments of the present application can be integrated in one processing module, or each module or sub-module can exist physically, or two or more modules or sub-modules can be integrated in one module. The integrated module or sub-module can be realized in the form of hardware or in the form of a software functional module or sub-module.

[0214] The skilled person can further realize that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be realized in electronic hardware, computer software, or a combination of both. In order to clearly illustrate the interchangeability of hardware and software, the components and steps of each example have been described in general terms in the above description. Whether the functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. The skilled person can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of the present application.

[0215] The steps of a method or algorithm described in connection with the embodiments disclosed herein can be embodied directly in hardware, in a software executed by a processor, or in a combination of the two. A software unit can reside in RAM (random access memory), flash memory, ROM (read-only memory), EPROM (erasable programmable ROM), EEPROM (electrically erasable programmable ROM), registers, hard disk, a removable disk, a CD-ROM, or any other form of storage medium known in the art.

[0216] Finally, it should be noted that the terms "first", "second", and the like, herein do not denote any order, quantity, combination, or importance, but rather are used to distinguish one element from another, and are more especially used for the purpose of identification in claims.

[0217] The foregoing description of the disclosed embodiments enables a person skilled in the art to implement or use the application. Numerous modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein can be applied to other embodiments without departing from the spirit or scope of the application. Therefore, the present application is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method of communication emitter identification, the method comprising: receiving a signal from a communication emitter; determining a signal characteristic of the signal; and identifying the communication emitter based on the signal characteristic. The method comprises the following steps: determining a background signal in the acquired radiation source signal, and acquiring signal features of the radiation source signal and signal features of the background signal; the background signal comprises signal components other than a central signal component of the radiation source signal, and the central signal component comprises signal components within a set bandwidth range based on a central frequency point of the radiation source signal; filtering the signal features of the background signal from the signal features of the radiation source signal to obtain individual fingerprint features of the radiation source signal; identifying a communication radiation source corresponding to the radiation source signal according to the individual fingerprint features.

2. The method of claim 1, wherein, The step of determining the background signal in the acquired radiation source signal comprises: extracting a signal of a set bandwidth from the acquired radiation source signal based on a central frequency point of the radiation source signal to obtain a central narrowband signal; acquiring a background signal from a signal within a target frequency range of the radiation source signal, wherein the target frequency range is a frequency range outside a frequency range in which the central narrowband signal is located.

3. The method of claim 2, wherein, The frequency range in which the background signal is located is adjacent to the frequency range in which the central narrowband signal is located.

4. The method of claim 1, wherein, The acquisition process of the radiation source signal comprises: receiving a high-frequency signal transmitted by a communication radiation source; converting the high-frequency signal into an intermediate-frequency signal, and taking the intermediate-frequency signal as the radiation source signal.

5. The method of claim 1, wherein, The step of acquiring the signal features of the radiation source signal and the signal features of the background signal comprises: calculating high-order statistics of the radiation source signal and high-order statistics of the background signal based on a Gaussian mixture model; taking the high-order statistics of the radiation source signal as the signal features of the radiation source signal, and taking the high-order statistics of the background signal as the signal features of the background signal.

6. The method of claim 5, wherein, The step of calculating the high-order statistics of the radiation source signal and the high-order statistics of the background signal based on the Gaussian mixture model comprises: establishing a signal Gaussian mixture model corresponding to the radiation source signal and a signal Gaussian mixture model corresponding to the background signal based on the Gaussian mixture model and the radiation source signal and the background signal; summing signal Gaussian mixture model parameters corresponding to each signal frame of the radiation source signal according to the signal Gaussian mixture model corresponding to the radiation source signal to obtain the high-order statistics of the radiation source signal; summing signal Gaussian mixture model parameters corresponding to each signal frame of the background signal according to the signal Gaussian mixture model corresponding to the background signal to obtain the high-order statistics of the background signal; wherein the signal Gaussian mixture model parameters corresponding to a signal frame comprise Gaussian function weights, class center mean values and standard deviation vectors; the summing of the signal Gaussian mixture model parameters comprises summing the Gaussian function weights corresponding to all signal frames, summing the class center mean values corresponding to all signal frames, and summing the standard deviation vectors corresponding to all signal frames; and the high-order statistics are a vector group composed of the sum of the Gaussian function weights, the sum of the class center mean values and the sum of the standard deviation vectors.

7. The method of claim 6, wherein, The establishment process of the signal Gaussian mixture model comprises: The Gaussian mixture model is iteratively trained by sequentially inputting each signal frame in the input signal into the Gaussian mixture model, so that the Gaussian mixture model divides each signal frame in the input signal into at least one signal frame subset, and a signal Gaussian mixture model corresponding to the input signal is obtained; According to the class center mean calculation rule, Gaussian function weight calculation rule and standard deviation vector calculation rule of the Gaussian mixture model, the class center mean, Gaussian function weight and standard deviation vector of each signal frame subset are calculated; According to the class center mean, Gaussian function weight and standard deviation vector of each signal frame subset in the input signal, the signal Gaussian mixture model parameters corresponding to the input signal are determined; The input signal includes a radiation source signal or a background signal.

8. The method of claim 1, wherein, The signal characteristics of the radiation source signal include high-order statistics of the radiation source signal, and the signal characteristics of the background signal include high-order statistics of the background signal; The individual fingerprint characteristics of the radiation source signal are obtained by filtering the signal characteristics of the background signal from the signal characteristics of the radiation source signal, including: The high-order statistics of the radiation source signal and the high-order statistics of the background signal are subjected to variance calculation to obtain at least one group of variance vectors; A group of variance vectors is selected from all the variance vectors as the individual fingerprint characteristics of the radiation source signal.

9. The method of claim 8, wherein, The group of variance vectors is selected from all the variance vectors as the individual fingerprint characteristics of the radiation source signal, including: The smallest group of variance vectors is selected from all the variance vectors as the individual fingerprint characteristics of the radiation source signal.

10. The method of claim 1, wherein, According to the individual fingerprint characteristics, the communication radiation source corresponding to the radiation source signal is identified, including: The individual fingerprint characteristics are input into a pre-trained feature classification model to identify the communication radiation source corresponding to the radiation source signal. The feature classification model is a model obtained by training a neural network using radiation source individual characteristics carrying radiation source identifiers.

11. A communication radiation source identification apparatus, characterized by, Including: An acquisition module is configured to acquire a background signal in a radiation source signal and acquire signal characteristics of the radiation source signal and signal characteristics of the background signal; the background signal includes signal components other than central signal components of the radiation source signal, and the central signal components include signal components within a set bandwidth range based on a center frequency point of the radiation source signal; A feature filtering module is configured to filter the signal characteristics of the background signal from the signal characteristics of the radiation source signal to obtain individual fingerprint characteristics of the radiation source signal; A radiation source identification module is configured to identify a communication radiation source corresponding to the radiation source signal according to the individual fingerprint characteristics.

12. A communication radiation source identification device, comprising: Including: A memory and a processor; The memory is connected with the processor and is configured to store programs; The processor is configured to realize the communication radiation source identification method in any one of claims 1 to 10 by running the programs in the memory.

13. A storage medium, characterized by The storage medium has a computer program stored thereon, and the computer program is executed by a processor to realize the communication radiation source identification method in any one of claims 1 to 10.

Images