Interference signal identification method and device based on neural network, medium and equipment

Abstract

The invention discloses an interference signal identification method and device based on a neural network, a medium and equipment, and belongs to the technical field of communication. Acquiring a to-be-identified signal, wherein the to-be-identified signal is obtained by superposing a noise signal, a useful signal and interference signals of n interference types; obtaining a neural network, wherein the neural network comprises a three-layer convolution attention module, a multi-branch fusion module, a double-branch hybrid parallel attention module and a full connection layer which are connected in parallel; converting the signal to be identified into a corresponding time-frequency image; and processing the time-frequency image by using a module in the neural network to obtain n interference types of the interference signal in the to-be-identified signal. According to the method, the model structure can be simplified, the training speed can be improved, the correlation of data in a larger spatial range can be captured, the recognition accuracy can be improved, the classification performance and generalization ability of the model can be improved, feature optimization selection can be realized, the features of signals of different modes can be considered in the recognition process, and the recognition accuracy can be improved.

Description

Technical Field

[0001] The present application relates to the field of communication technology, and in particular to a method, device, medium and equipment for identifying interference signals based on a neural network. Background Art

[0002] In modern wireless communications, with the rapid development of technology and the widespread expansion of its application, communication systems are increasingly susceptible to various interferences. These interferences come from a variety of sources, including natural environmental noise, electromagnetic interference generated by equipment operation, and malicious interference. This interference not only degrades communication quality but can also cause communication interruptions, severely impacting the normal operation of commercial activities, security monitoring, and emergency rescue operations. Therefore, anti-interference technology has become a crucial technical means to ensure the stable operation of wireless communication systems.

[0003] The prerequisite for anti-interference is the accurate and real-time classification and identification of interference signals in received signals. As a key component of wireless communication anti-interference, interference signal identification supports subsequent interference elimination and suppression methods, improving the performance of satellite communication systems.

[0004] Existing interference signal recognition methods typically use different networks for different modal data. For example, Model 1 is used for time domain data, and Model 2 is used for frequency domain data. However, this method only uses a single modal data during the recognition process, ignoring the features in other modal data, resulting in low recognition accuracy. Summary of the Invention

[0005] This application provides a neural network-based interference signal identification method, device, medium, and equipment to address the problem of low interference signal identification accuracy in related technologies. The technical solution is as follows:

[0006] According to a first aspect of the present application, a method for identifying interference signals based on a neural network is provided, the method comprising:

[0007] Acquire a signal to be identified, where the signal to be identified is obtained by superimposing at least a noise signal, a useful signal, and interference signals of n interference types, where n is an unknown number greater than or equal to 1;

[0008] Obtain a pre-trained neural network, wherein the neural network includes at least a three-layer convolutional attention module in parallel, a multi-branch fusion module, a dual-branch hybrid parallel attention module, and a fully connected layer, the convolutional attention module includes at least a two-dimensional depth convolution module, a channel attention module, a two-dimensional point-by-point convolution module, and a spatial attention module, the multi-branch fusion module includes at least a two-dimensional convolution module, a compression and excitation attention module, and an h-swish activation function, and the dual-branch hybrid parallel attention module includes at least a parallel multi-head self-attention module, a convolutional block attention module, a one-dimensional layer normalization layer, and a feedforward neural network;

[0009] Converting the signal to be identified into a corresponding time-frequency image;

[0010] The time-frequency image is processed using a module in the neural network to obtain n interference types of interference signals in the signal to be identified.

[0011] In one possible implementation, when the neural network includes three layers of convolutional attention modules connected in parallel, processing the time-frequency image using each module in the neural network includes:

[0012] Performing a convolution operation on the time-frequency image using the two-dimensional depth convolution module to obtain a first intermediate result;

[0013] Performing a channel attention calculation on the first intermediate result using the channel attention module to obtain a second intermediate result;

[0014] performing a convolution operation on the second intermediate result using the two-dimensional point-by-point convolution module to obtain a third intermediate result;

[0015] The spatial attention module is used to perform a spatial attention operation on the third intermediate result to obtain an output result of the convolutional attention module.

[0016] In a possible implementation, when the neural network includes a multi-branch fusion module and other modules, processing the time-frequency image using the modules in the neural network includes:

[0017] The output results of the three-layer convolutional attention module in parallel are concatenated into a feature map;

[0018] Performing a convolution operation on the feature map using the two-dimensional convolution module to obtain a fourth intermediate result;

[0019] performing an attention operation on the fourth intermediate result using the compression and stimulation attention module to obtain a fifth intermediate result;

[0020] Using the h-swish activation function to operate on the fifth intermediate result to obtain a sixth intermediate result;

[0021] The sixth intermediate result is processed by other modules to obtain output results of the other modules.

[0022] In one possible implementation, when the neural network includes a dual-branch hybrid parallel attention module and a fully connected layer, processing the time-frequency image using the module in the neural network includes:

[0023] Normalizing the output results of the other modules using the one-dimensional normalization layer in the dual-branch hybrid parallel attention module, multiplying the normalized results by three different weight matrices respectively to obtain a query, a key, and a value, performing an attention operation and concatenating the query, the key, and the value using the multi-head self-attention module in the dual-branch hybrid parallel attention module to obtain a seventh intermediate result;

[0024] Using the channel attention module in the convolution block attention module to perform a channel attention operation on the output results of the other modules, and using the spatial attention module in the convolution block attention module to perform a spatial attention operation on the operation results, to obtain an eighth intermediate result;

[0025] performing a normalization operation on the sum of the seventh intermediate result and the eighth intermediate result using the one-dimensional layer normalization layer to obtain a ninth intermediate result;

[0026] Processing the ninth intermediate result using the feedforward neural network to obtain an output result of the dual-branch hybrid parallel attention module;

[0027] The fully connected layer is used to calculate the output result of the dual-branch hybrid parallel attention module to obtain n interference types of the interference signal in the signal to be identified.

[0028] In a possible implementation, converting the interference signal into a corresponding time-frequency image includes:

[0029] Converting the signal to be identified into a corresponding time-frequency image based on short-time Fourier transform; or

[0030] The signal to be identified is converted into a corresponding time-frequency image based on wavelet transform.

[0031] In a possible implementation, the interference type includes at least one of single-tone interference, multi-tone interference, linear frequency sweep interference, pulse interference, comb spectrum interference, narrowband noise interference, and broadband noise interference.

[0032] In a possible implementation, the method further includes:

[0033] Acquire multiple groups of training samples, each group of training samples including interference signal samples and pre-labeled labels, wherein the interference signal samples are obtained by superimposing at least a noise signal, a useful signal, and an interference signal of at least one interference type, and the labels are used to label all actual interference types corresponding to the interference signal samples;

[0034] creating the neural network;

[0035] For each set of training samples, using the modules in the neural network to process the interference signal samples in the training samples to obtain the corresponding predicted interference type; calculating the loss value according to the predicted interference type, the actual interference type and a preset loss function;

[0036] The parameters of the neural network are adjusted according to the loss value, and the training is stopped after the parameters meet the preset conditions.

[0037] According to a second aspect of the present application, a neural network-based interference signal identification device is provided, the device comprising:

[0038] An acquisition module is used to acquire a signal to be identified, where the signal to be identified is obtained by superimposing at least a noise signal, a useful signal, and interference signals of n interference types, where n is an unknown number greater than or equal to 1;

[0039] The acquisition module is further used to acquire a pre-trained neural network, which includes at least a three-layer convolutional attention module, a multi-branch fusion module, a dual-branch hybrid parallel attention module and a fully connected layer in parallel, the convolutional attention module includes at least a two-dimensional depth convolution module, a channel attention module, a two-dimensional point-by-point convolution module and a spatial attention module, the multi-branch fusion module includes at least a two-dimensional convolution module, a compression and excitation attention module and an h-swish activation function, and the dual-branch hybrid parallel attention module includes at least a parallel multi-head self-attention module, a convolutional block attention module, a one-dimensional layer normalization layer and a feedforward neural network;

[0040] A conversion module, configured to convert the signal to be identified into a corresponding time-frequency image;

[0041] An identification module is used to process the time-frequency image using a module in the neural network to obtain n interference types of interference signals in the signal to be identified.

[0042] According to a third aspect of the present application, a computer-readable storage medium is provided, in which at least one instruction is stored. The at least one instruction is loaded and executed by a processor to implement the interference signal identification method based on a neural network as described above.

[0043] According to a fourth aspect of the present application, a computer device is provided, which includes the above-mentioned neural network-based interference signal identification device.

[0044] The beneficial effects of the technical solution provided by this application include at least:

[0045] Since the DSC (Depthwise Separable Convolution) module includes a two-dimensional depth convolution module and a two-dimensional point-by-point convolution module, by improving the DSC module, a channel attention module is added after the two-dimensional depth convolution module and a spatial attention module is added after the two-dimensional point-by-point convolution module, so as to improve the DSC module into a convolution attention module. This can simplify the model structure while maintaining performance, thereby improving the training speed, and can capture the correlation of data in a larger spatial range, thereby improving recognition accuracy.

[0046] The dual-branch hybrid parallel attention module includes a parallel multi-head self-attention module and a convolutional block attention module. The multi-head self-attention module can extract long-distance dependencies and enhance the information connection between features. The channel attention module and spatial attention module in the convolutional block attention module can strengthen effective feature information and weaken invalid feature information, thereby improving the classification performance and generalization ability of the model and realizing feature optimization selection.

[0047] The signal to be identified includes interference signals of n interference types. When the value of n is different, the neural network can identify multi-modal signals to be identified and consider the characteristics of signals to be identified of different modes during the recognition process, thereby improving the recognition accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0049] Figure 1 is a schematic diagram of the structure of a neural network according to some exemplary embodiments;

[0050] Figure 2 is a structural diagram of a convolutional attention module according to some exemplary embodiments;

[0051] Figure 3 is a structural diagram of a dual-branch hybrid parallel attention module according to some exemplary embodiments;

[0052] Figure 4This is a flowchart of a neural network-based training method provided by one embodiment of the present application;

[0053] Figure 5 This is a flowchart of a neural network-based interference signal identification method provided by one embodiment of the present application;

[0054] Figure 6 This is a flowchart of a neural network-based interference signal identification method provided by one embodiment of the present application;

[0055] Figure 7 This is a structural block diagram of an interference signal identification device based on a neural network provided by an embodiment of the present application;

[0056] Figure 8 This is a structural block diagram of a computer device provided in one embodiment of the present application. DETAILED DESCRIPTION

[0057] In order to make the objectives, technical solutions and advantages of the embodiments of the present application clearer, the implementation methods of the present application will be further described in detail below with reference to the accompanying drawings.

[0058] In this application, it is necessary to pre-create a neural network and train the neural network so that the trained neural network can identify multimodal interference signals. The structure and training process of the neural network are described below.

[0059] (1) Structure of neural network

[0060] The neural network in this application includes at least a three-layer convolutional attention module (DCPS) in parallel, a multi-branch fusion module (MFFU), a dual-branch hybrid parallel attention module (HPA) and a fully connected layer (FC), and may also include a maximum pooling layer (MaxPooling), a fusion module (FB), an average pooling layer (Avg Pooling), a block embedding (PatchEmbedding) and a vertical strip embedding (Vertical Strip Embedding).

[0061] In one example, the connection relationship between the modules in the neural network is as follows: Figure 1 The neural network consists of three parallel layers of convolutional attention modules, a multi-branch fusion module, a maximum pooling layer, a convolutional attention module, a fusion module, an average pooling layer, a convolutional attention module, a fusion module, an average pooling layer, a convolutional attention module, a fusion module, a parallel block embedding and vertical strip embedding, a dual-branch hybrid parallel attention module, and two fully connected layers.

[0062] In an example, the structure of a convolutional attention module is as follows Figure 2As shown. A convolutional attention module includes at least a two-dimensional depth convolution module (DConv), a channel attention module (CA), a two-dimensional point-by-point convolution module (PConv), and a spatial attention module (SA). Among them, the convolution kernel of the two-dimensional depth convolution module is k×k, and the value of k in different convolutional attention modules is different. For example, the neural network includes three layers of parallel convolutional attention modules, and the values ​​of k are 3, 5, and 7 respectively. Then the convolution kernels of the three two-dimensional depth convolution modules in the three-layer convolutional attention module are 3×3, 5×5, and 7×7 respectively. The convolution kernel of the two-dimensional point-by-point convolution module is 1×1.

[0063] In one example, the multi-branch fusion module includes at least a two-dimensional convolution module, a compression and excitation attention module, and an h-swish activation function.

[0064] In one example, the structure of the dual-branch hybrid parallel attention module is as follows Figure 3 As shown in Figure 2, the dual-branch hybrid parallel attention module includes at least a parallel multi-head self-attention module, a convolutional block attention module, a one-dimensional layer normalization layer, and a feedforward neural network. The parallel multi-head self-attention module includes at least a one-dimensional layer normalization layer (Norm) and a multi-head self-attention module (MSA), the convolutional block attention module includes at least a series-connected channel attention module (CA) and a spatial attention module (SA), and the feedforward neural network includes at least two fully connected layers.

[0065] (2) Neural network training process

[0066] like Figure 4 As shown in Figure 2, the training process of a neural network includes the following steps:

[0067] Step 401: Acquire multiple groups of training samples, each group of training samples includes interference signal samples and pre-labeled labels. The interference signal samples are obtained by superimposing at least a noise signal, a useful signal, and an interference signal of at least one interference type. The labels are used to label all actual interference types corresponding to the interference signal samples.

[0068] Each set of training samples includes an interference signal sample and a label. The interference signal sample can be a single interference signal sample or a composite interference signal sample. A single interference signal is obtained by superimposing a noise signal, a useful signal, and an interference signal of any interference category. A composite interference signal is obtained by superimposing a noise signal, a useful signal, and interference signals of at least two interference categories.

[0069] Interference signal samples can be generated using simulation software, such as MATLAB software. One interference category corresponds to multiple interference signal samples with different interference-to-signal ratios. Interference signal samples are labeled with their interference categories. For example, a single interference signal sample is labeled with its corresponding single interference category, and a composite interference signal sample is labeled with its corresponding at least two interference categories.

[0070] In one embodiment, taking wireless communication as an example, the useful signal is a BPSK (Binary Phase Shift Keying) modulated signal, the noise signal is additive white Gaussian noise, and the signal parameters are set as follows: signal-to-noise ratio is 5dB, interference-to-signal ratio is [-30dB, 10dB], interval is 4dB, carrier frequency is 140KHz, and sampling frequency is 800KHz.

[0071] The interference type includes at least one of single-tone interference, multi-tone interference, linear frequency sweep interference, pulse interference, comb spectrum interference, narrowband noise interference and broadband noise interference.

[0072] Among them, the frequency point of single-tone interference is randomly generated within the bandwidth, the frequency set of multi-tone interference is randomly generated within the bandwidth, the initial frequency of linear sweep interference is 120KHz, the sweep rate is 8MHz, the pulse interference is a rectangular pulse with a duty cycle of 2%, the total bandwidth of comb spectrum interference accounts for more than 1 / 2, the bandwidth of narrowband noise interference accounts for 1 / 10, and the bandwidth of broadband noise interference accounts for more than 2 / 3.

[0073] The number of signal samples for each interference type and each signal-to-interference ratio is 250, and they are divided into training, validation, and test sets in a ratio of 6:2:2. The training set is the training sample set, while the validation and test sets are used to verify whether the parameters of the neural network meet the preset conditions.

[0074] Step 402: Create a neural network.

[0075] Computer equipment can create Figure 2 The neural network shown.

[0076] Step 403: For each set of training samples, the interference signal samples in the training samples are processed using the modules in the neural network to obtain the corresponding predicted interference type; and the loss value is calculated according to the predicted interference type, the actual interference type and the preset loss function.

[0077] The computer device can convert the interference signal samples in each group of training samples into corresponding time-frequency images, and then input the time-frequency images into the neural network. The neural network processes the time-frequency images and outputs the predicted interference type, and calculates the loss value based on the predicted interference type, the actual interference type and the preset loss function.

[0078] When converting the interference signal samples into a time-frequency image, the computer device can convert the signal to be identified into a corresponding time-frequency image based on a short-time Fourier transform; or convert the signal to be identified into a corresponding time-frequency image based on a wavelet transform. After converting the interference signal samples into a time-frequency image, the advantages of both time-domain and frequency-domain features can be combined, thereby improving recognition accuracy.

[0079] Step 404: Adjust the parameters of the neural network according to the loss value until the parameters meet the preset conditions and then stop training.

[0080] During training, the computer uses the Adam optimizer, an adaptive optimization algorithm, to adaptively adjust the learning rate based on the loss value calculated by the loss function. The initial learning rate is 0.001, and the training is repeated 50 times. After the training stabilizes, the validation and test sets are fed into the trained neural network to test the accuracy of interference recognition.

[0081] like Figure 5 As shown, it shows a method flow chart of a neural network-based interference signal identification method provided by an embodiment of the present application. The neural network-based interference signal identification method can be applied to computer equipment. The neural network-based interference signal identification method may include:

[0082] Step 501: Acquire a signal to be identified, where the signal to be identified is obtained by superimposing at least a noise signal, a useful signal, and interference signals of n interference types, where n is an unknown number greater than or equal to 1.

[0083] Here, n is an unknown number greater than or equal to 1, that is, the signal to be identified may include interference signals of one interference type, or may include interference signals of at least two interference types.

[0084] Step 502: Obtain a pre-trained neural network.

[0085] The computer device obtains the neural network trained through the above training process.

[0086] Step 503: convert the signal to be identified into a corresponding time-frequency image.

[0087] Step 504: Process the time-frequency image using the modules in the neural network to obtain n interference types of the interference signal in the signal to be identified.

[0088] The computer device inputs the time-frequency image into the neural network. Each module in the neural network processes the time-frequency image and outputs a predicted interference type, which is used to identify n interference types of the interference signal in the signal to be identified.

[0089] To sum up, the interference signal recognition method based on neural network provided in the embodiment of the present application, since the DSC module includes a two-dimensional deep convolution module and a two-dimensional point-by-point convolution module, by improving the DSC module, adding a channel attention module after the two-dimensional deep convolution module and adding a spatial attention module after the two-dimensional point-by-point convolution module, so as to improve the DSC module into a convolution attention module, can simplify the model structure while maintaining performance, thereby improving the training speed, and can capture the correlation of data in a larger spatial range, thereby improving the recognition accuracy.

[0090] The dual-branch hybrid parallel attention module includes a parallel multi-head self-attention module and a convolutional block attention module. The multi-head self-attention module can extract long-distance dependencies and enhance the information connection between features. The channel attention module and spatial attention module in the convolutional block attention module can strengthen effective feature information and weaken invalid feature information, thereby improving the classification performance and generalization ability of the model and realizing feature optimization selection.

[0091] The signal to be identified includes interference signals of n interference types. When the value of n is different, the neural network can identify multi-modal signals to be identified and consider the characteristics of signals to be identified of different modes during the recognition process, thereby improving the recognition accuracy.

[0092] like Figure 6 As shown, it shows a flowchart of an interference signal identification method based on a neural network provided by an embodiment of the present application. The interference signal identification method based on a neural network can be applied to a computer device. The interference signal identification method based on a neural network may include:

[0093] Step 601: Acquire a signal to be identified, where the signal is obtained by superimposing at least a noise signal, a useful signal, and interference signals of n interference types, where n is an unknown number greater than or equal to 1.

[0094] Where n is an unknown number greater than or equal to 1, that is, the signal to be identified may contain interference signals of one interference type or may contain interference signals of at least two interference types. The interference types include at least one of single-tone interference, multi-tone interference, linear frequency sweep interference, pulse interference, comb spectrum interference, narrowband noise interference, and broadband noise interference.

[0095] Step 602: Obtain a pre-trained neural network.

[0096] The computer device obtains the neural network trained through the above training process.

[0097] Step 603: Convert the signal to be identified into a corresponding time-frequency image based on short-time Fourier transform; or convert the signal to be identified into a corresponding time-frequency image based on wavelet transform.

[0098] In one example, the resolution of the time-frequency image is 3×224×224.

[0099] After converting the interference signal samples into time-frequency images, the advantages of time domain features and frequency domain features can be concentrated, thereby improving the recognition accuracy.

[0100] Step 604: Process the time-frequency image using the modules in the neural network to obtain n interference types of the interference signal in the signal to be identified.

[0101] When the neural network includes three layers of convolutional attention modules in parallel, processing the time-frequency image using each module in the neural network can include: using a two-dimensional depth convolution module to perform a convolution operation on the time-frequency image to obtain a first intermediate result; using a channel attention module to perform a channel attention calculation on the first intermediate result to obtain a second intermediate result; using a two-dimensional point-by-point convolution module to perform a convolution operation on the second intermediate result to obtain a third intermediate result; using a spatial attention module to perform a spatial attention operation on the third intermediate result to obtain the output result of the convolution attention module.

[0102] The input of the convolutional attention module is T∈R C×H×W , first pass through the three-layer two-dimensional depth convolution module, the convolution kernel size is k×k, and get the first intermediate result T1∈R C×H×W ; Input the first intermediate result into the channel attention module to obtain the second intermediate result T2∈R C×H×W ; Then the second intermediate result is input into the two-dimensional point-by-point convolution module and the spatial attention module to obtain the output result T3∈R with channel and spatial attention C0×H×W The whole process can be summarized as follows:

[0103] DCPS(T,k)=SA(PConv(CA(DConv(T,k)))) (1)

[0104] Among them, T represents the feature map (time-frequency image) of the input convolution attention module, k represents the size of the convolution kernel in the two-dimensional depth convolution module, C represents the number of channels, H represents the height of the feature map, W represents the width of the feature map, T1 represents the feature map after two-dimensional depth convolution processing, T2 represents the feature map after channel attention mechanism processing, and T3 represents the feature map after point-by-point convolution and spatial attention mechanism processing.

[0105] The meaning of formula (1) is that the feature map T undergoes two-dimensional depth convolution with a convolution kernel of k×k, channel attention, two-dimensional point-by-point convolution and spatial attention in sequence.

[0106] The two-dimensional depth convolution module and the two-dimensional point-by-point convolution module are used to reduce the number of model parameters to reduce the model calculation complexity.

[0107] The channel attention module focuses on the relationship between different channels of the time-frequency image, assigns different weights to each channel, and strengthens the focus on important channels, which helps to improve the representation ability of the model.

[0108] The spatial attention module focuses on the spatial structure within the time-frequency image and assigns different weights to each pixel position, enabling the model to focus on important areas in the time-frequency image.

[0109] When the neural network includes multi-branch fusion modules and other modules, processing the time-frequency image using the modules in the neural network can include: splicing the output results of the three-layer convolution attention modules in parallel into a feature map; using the two-dimensional convolution module to perform a convolution operation on the feature map to obtain a fourth intermediate result; using the compression and excitation attention module to perform an attention operation on the fourth intermediate result to obtain a fifth intermediate result; using the h-swish activation function to operate on the fifth intermediate result to obtain a sixth intermediate result; using other modules to process the sixth intermediate result to obtain the output results of other modules.

[0110] The outputs of the three-layer convolutional attention module are F1, F2, and F3, each of size 3 × H × W. These three outputs are concatenated along the channel dimension to produce a 9 × H × W feature map that combines the feature information of each convolutional attention module. A 1 × 1 two-dimensional convolutional layer is then used to obtain a preliminary feature fusion result, and the number of channels is adjusted to obtain the fourth intermediate result. Finally, the sixth intermediate result is obtained by performing calculations using compressed and stimulated attention and the h-swish activation function. The entire process can be summarized as follows:

[0111]

[0112] Among them, F1, F2 and F3 represent the output results of the three-layer convolutional attention module respectively, SE represents the compression and excitation attention module, HS represents the h-swish activation function, and concat represents the concatenation operation of the three output results.

[0113] After obtaining the sixth intermediate result through the multi-branch fusion module, it is necessary to input the sixth intermediate result into other modules for processing, and finally obtain the input feature map of the dual-branch hybrid parallel attention module. Figure 2 As shown, other modules include max pooling layer, convolutional attention module, fusion module, average pooling layer, convolutional attention module, fusion module, average pooling layer, convolutional attention module, fusion module, parallel block embedding and vertical strip embedding.

[0114] When the neural network includes a dual-branch hybrid parallel attention module and a fully connected layer, processing the time-frequency image using the modules in the neural network may include: using the one-dimensional layer normalization layer in the dual-branch hybrid parallel attention module to normalize the output results of other modules, multiplying the normalized results with three different weight matrices respectively to obtain the query, key and value, using the multi-head self-attention module in the dual-branch hybrid parallel attention module to perform attention operations and splicing on the query, key and value to obtain the seventh intermediate result; using the channel attention module in the convolution block attention module to perform channel attention operations on the output results of other modules, and using the spatial attention module in the convolution block attention module to perform spatial attention operations on the operation results to obtain the eighth intermediate result; using the one-dimensional layer normalization layer to normalize the sum of the seventh intermediate result and the eighth intermediate result to obtain the ninth intermediate result; using the feedforward neural network to process the ninth intermediate result to obtain the output result of the dual-branch hybrid parallel attention module; using the fully connected layer to calculate the output result of the dual-branch hybrid parallel attention module to obtain n interference types of the interference signal in the signal to be identified.

[0115] The input feature map is X∈R C×M , first process the input feature map through the multi-head self-attention module and the convolution block attention module respectively to obtain the seventh intermediate result and the eighth intermediate result; then, the seventh intermediate result and the eighth intermediate result are summed and processed through the one-dimensional layer normalization layer and the feedforward neural network to obtain the output result X1∈R of the dual-branch hybrid parallel attention module C×M , M represents the length of the feature sequence. The whole process can be summarized as:

[0116] X'=X MSA +X CBAM (3)

[0117] X'=FFN(LN(X'))+X' (4)

[0118] Among them, X MSA Characterizing the multi-head self-attention module, X CBAM represents the convolutional block attention module, LN represents the one-dimensional layer normalization layer, and FFN represents the feedforward neural network.

[0119] For the multi-head self-attention module, the input feature map is first processed through a one-dimensional normalization layer, and then the normalized result is multiplied by different weight matrices to obtain the query, key, and value, which are input into the multi-head self-attention module to obtain the seventh intermediate result. The whole process can be summarized as follows:

[0120]

[0121] head i=Attention(QW i Q ,KW i K ,VW i V ) (6)

[0122] MSA=concat(head1,head2,…,head n )W (7)

[0123] Among them, Q represents the query, K represents the key, V represents the value, d k Represents the dimension of the key, softmax represents the normalized exponential function, W represents the weight matrix, and concat represents the concatenation operation.

[0124] The convolutional block attention module is composed of a channel attention module and a spatial attention module in series. The input of the channel attention module is the input feature map F∈R C×M , the output is a one-dimensional channel attention feature map M C ∈R C×1 , input the one-dimensional channel attention feature map into the spatial attention module to obtain a one-dimensional spatial attention feature map M S ∈R 1×M , which is the eighth intermediate result. The whole process can be summarized as:

[0125]

[0126]

[0127] Among them, F represents the feature map of the input channel attention module, R represents the real number field, C represents the number of channels, H represents the height of the feature map, W represents the width of the feature map, F' represents the feature sequence after processing by the channel attention mechanism, and F" represents the feature sequence after processing by the spatial attention mechanism.

[0128] Formula (8) means: the feature sequence after processing by the channel attention mechanism = channel attention feature sequence × original feature sequence.

[0129] Formula (9) means: feature sequence processed by spatial attention mechanism = spatial attention feature sequence × feature sequence processed by channel attention mechanism.

[0130] The fully connected layer maps the outputs of multiple neurons to probabilities in [0, 1] through the softmax activation function to achieve multi-label classification, and the loss function is the cross entropy loss function.

[0131] To sum up, the interference signal recognition method based on neural network provided in the embodiment of the present application, since the DSC module includes a two-dimensional deep convolution module and a two-dimensional point-by-point convolution module, by improving the DSC module, adding a channel attention module after the two-dimensional deep convolution module and adding a spatial attention module after the two-dimensional point-by-point convolution module, so as to improve the DSC module into a convolution attention module, can simplify the model structure while maintaining performance, thereby improving the training speed, and can capture the correlation of data in a larger spatial range, thereby improving the recognition accuracy.

[0132] The dual-branch hybrid parallel attention module includes a parallel multi-head self-attention module and a convolutional block attention module. The multi-head self-attention module can extract long-distance dependencies and enhance the information connection between features. The channel attention module and spatial attention module in the convolutional block attention module can strengthen effective feature information and weaken invalid feature information, thereby improving the classification performance and generalization ability of the model and realizing feature optimization selection.

[0133] The signal to be identified includes interference signals of n interference types. When the value of n is different, the neural network can identify multi-modal signals to be identified and consider the characteristics of signals to be identified of different modes during the recognition process, thereby improving the recognition accuracy.

[0134] like Figure 7 As shown, it shows a structural block diagram of an interference signal identification device based on a neural network provided by an embodiment of the present application. The interference signal identification device based on a neural network can be applied to a computer device. The interference signal identification device based on a neural network can include:

[0135] An acquisition module 710 is configured to acquire a signal to be identified, where the signal to be identified is obtained by superimposing at least a noise signal, a useful signal, and interference signals of n interference types, where n is an unknown number greater than or equal to 1;

[0136] The acquisition module 710 is further used to obtain a pre-trained neural network, which includes at least a three-layer convolutional attention module, a multi-branch fusion module, a dual-branch hybrid parallel attention module and a fully connected layer in parallel. The convolutional attention module includes at least a two-dimensional depth convolution module, a channel attention module, a two-dimensional point-by-point convolution module and a spatial attention module. The multi-branch fusion module includes at least a two-dimensional convolution module, a compression and excitation attention module and an h-swish activation function. The dual-branch hybrid parallel attention module includes at least a parallel multi-head self-attention module, a convolutional block attention module, a one-dimensional layer normalization layer and a feedforward neural network.

[0137] A conversion module 720 is used to convert the signal to be identified into a corresponding time-frequency image;

[0138] The identification module 730 is used to process the time-frequency image using the modules in the neural network to obtain n interference types of the interference signal in the signal to be identified.

[0139] In an optional embodiment, when the neural network includes three layers of convolutional attention modules connected in parallel, the recognition module 730 is further configured to:

[0140] Performing a convolution operation on the time-frequency image using a two-dimensional depth convolution module to obtain a first intermediate result;

[0141] Use the channel attention module to perform channel attention calculation on the first intermediate result to obtain a second intermediate result;

[0142] Performing a convolution operation on the second intermediate result using a two-dimensional point-by-point convolution module to obtain a third intermediate result;

[0143] The spatial attention module is used to perform spatial attention operation on the third intermediate result to obtain the output result of the convolutional attention module.

[0144] In an optional embodiment, when the neural network includes a multi-branch fusion module and other modules, the identification module 730 is further configured to:

[0145] The output results of the three-layer convolutional attention module in parallel are concatenated into a feature map;

[0146] Use the two-dimensional convolution module to perform a convolution operation on the feature map to obtain a fourth intermediate result;

[0147] performing an attention operation on the fourth intermediate result using the compression and incentive attention module to obtain a fifth intermediate result;

[0148] The fifth intermediate result is operated by using the h-swish activation function to obtain a sixth intermediate result;

[0149] The sixth intermediate result is processed by other modules to obtain output results of other modules.

[0150] In an optional embodiment, when the neural network includes a dual-branch hybrid parallel attention module and a fully connected layer, the recognition module 730 is further configured to:

[0151] The one-dimensional normalization layer in the dual-branch hybrid parallel attention module is used to normalize the output results of other modules. The normalized results are multiplied by three different weight matrices to obtain the query, key, and value. The multi-head self-attention module in the dual-branch hybrid parallel attention module is used to perform attention operation and splicing on the query, key, and value to obtain the seventh intermediate result.

[0152] Using the channel attention module in the convolution block attention module to perform channel attention operation on the output results of other modules, and using the spatial attention module in the convolution block attention module to perform spatial attention operation on the operation results to obtain an eighth intermediate result;

[0153] Performing a normalization operation on the sum of the seventh intermediate result and the eighth intermediate result using a one-dimensional layer normalization layer to obtain a ninth intermediate result;

[0154] The ninth intermediate result is processed using a feedforward neural network to obtain the output result of the dual-branch hybrid parallel attention module;

[0155] The fully connected layer is used to calculate the output results of the dual-branch hybrid parallel attention module to obtain n interference types of the interference signal in the signal to be identified.

[0156] In an optional embodiment, the conversion module 720 is further configured to:

[0157] Convert the signal to be identified into a corresponding time-frequency image based on short-time Fourier transform; or

[0158] The signal to be identified is converted into a corresponding time-frequency image based on wavelet transform.

[0159] In an optional embodiment, the interference type includes at least one of single-tone interference, multi-tone interference, linear frequency sweep interference, pulse interference, comb spectrum interference, narrowband noise interference and broadband noise interference.

[0160] In an optional embodiment, the device further includes a training module for:

[0161] Acquire multiple sets of training samples, each set of training samples includes interference signal samples and pre-labeled labels, the interference signal samples are obtained by superimposing at least a noise signal, a useful signal, and an interference signal of at least one interference type, and the labels are used to label all actual interference types corresponding to the interference signal samples;

[0162] Create a neural network;

[0163] For each set of training samples, the interference signal samples in the training samples are processed using the modules in the neural network to obtain the corresponding predicted interference type; the loss value is calculated based on the predicted interference type, the actual interference type and the preset loss function;

[0164] The parameters of the neural network are adjusted according to the loss value until the parameters meet the preset conditions and the training is stopped.

[0165] To sum up, the interference signal recognition device based on neural network provided by the embodiment of the present application, since the DSC module includes a two-dimensional deep convolution module and a two-dimensional point-by-point convolution module, by improving the DSC module, adding a channel attention module after the two-dimensional deep convolution module and adding a spatial attention module after the two-dimensional point-by-point convolution module, so as to improve the DSC module into a convolution attention module, can simplify the model structure while maintaining performance, thereby improving the training speed, and can capture the correlation of data in a larger spatial range, thereby improving the recognition accuracy.

[0166] The dual-branch hybrid parallel attention module includes a parallel multi-head self-attention module and a convolutional block attention module. The multi-head self-attention module can extract long-distance dependencies and enhance the information connection between features. The channel attention module and spatial attention module in the convolutional block attention module can strengthen effective feature information and weaken invalid feature information, thereby improving the classification performance and generalization ability of the model and realizing feature optimization selection.

[0167] The signal to be identified includes interference signals of n interference types. When the value of n is different, the neural network can identify multi-modal signals to be identified and consider the characteristics of signals to be identified of different modes during the recognition process, thereby improving the recognition accuracy.

[0168] like Figure 8 , which shows a schematic structural diagram of a computer device 800 suitable for implementing an embodiment of the present invention. Figure 8 The computer device 800 shown is only an example and should not limit the functions and scope of use of the embodiments of the present invention.

[0169] like Figure 8 As shown, the computer device 800 includes a central processing unit (CPU) 801, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 802 or a program loaded from a storage unit 808 into a random access memory (RAM) 803. Various programs and data required for the operation of the system 800 are also stored in the RAM 803. The CPU 801, ROM 802, and RAM 803 are connected to each other via a bus 804. An input / output (I / O) interface is also connected to the bus 804.

[0170] The following components are connected to the I / O interface 805: an input section 806 including a keyboard, a mouse, and the like; an output section 807 including devices such as a cathode ray tube (CRT), a liquid crystal display (LCD), and a speaker; a storage section 808 including a hard disk; and a communication section 809 including a network interface card such as a LAN card or a modem. The communication section 809 performs communication processing via a network such as the Internet. A drive 810 is also connected to the I / O interface 805 as needed. A removable medium 811, such as a magnetic disk, an optical disk, a magneto-optical disk, or a semiconductor memory, is installed in the drive 810 as needed, so that computer programs read therefrom can be installed into the storage section 808 as needed.

[0171] In particular, according to the embodiments disclosed in the present invention, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, the embodiments disclosed in the present invention include a computer program product comprising a computer program carried on a computer-readable medium, the computer program comprising program code for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from a network via the communication section 809, and / or installed from a removable medium 811. When the computer program is executed by the central processing unit (CPU) 801, the above-mentioned functions defined in the system of the present invention are performed.

[0172] It should be noted that the computer-readable medium described in the present invention can be a computer-readable signal medium or a computer-readable storage medium, or any combination thereof. A computer-readable storage medium can be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of computer-readable storage media can include, but are not limited to, an electrical connection having one or more conductors, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination thereof. In the present invention, a computer-readable storage medium can be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device. In the present invention, a computer-readable signal medium can include a data signal propagated in baseband or as part of a carrier wave, carrying computer-readable program code. This propagated data signal can take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. A computer-readable signal medium may also be any computer-readable medium other than a computer-readable storage medium that can transmit, propagate, or transport a program for use by or in conjunction with an instruction execution system, apparatus, or device. Program code embodied on a computer-readable medium may be transmitted using any suitable medium, including but not limited to wireless, wireline, optical fiber cable, RF, or any suitable combination thereof.

[0173] The flowcharts and block diagrams in the accompanying drawings illustrate the possible implementation architecture, functions and operations of the systems, methods and computer program products according to various embodiments of the present invention. In this regard, each box in the flowchart or block diagram can represent a module, program segment, or a part of code, and the above-mentioned module, program segment, or a part of code contains one or more executable instructions for implementing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in an order different from that marked in the accompanying drawings. For example, two boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram or flowchart, and the combination of boxes in the block diagram or flowchart, can be implemented with a dedicated hardware-based system that performs the specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0174] The modules involved in the embodiments of the present invention may be implemented in software or hardware. The modules described may also be provided in a processor. For example, they may be described as follows: a processor includes a sending module, an acquisition module, a determination module, and a first processing module. The names of these modules do not, in some cases, limit the modules themselves. For example, the sending module may also be described as a "module for sending a picture acquisition request to the connected server."

[0175] The above specific embodiments do not limit the scope of protection of the present invention. Those skilled in the art will appreciate that various modifications, combinations, sub-combinations, and substitutions may occur depending on design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention are intended to be included within the scope of protection of the present invention.

Claims

1. A method for identifying interference signals based on a neural network, characterized in that: The method comprises: Acquire a signal to be identified, where the signal to be identified is obtained by superimposing at least a noise signal, a useful signal, and interference signals of n interference types, where n is an unknown number greater than or equal to 1; Obtain a pre-trained neural network, wherein the neural network includes at least a three-layer convolutional attention module in parallel, a multi-branch fusion module, a dual-branch hybrid parallel attention module, and a fully connected layer, the convolutional attention module includes at least a two-dimensional depth convolution module, a channel attention module, a two-dimensional point-by-point convolution module, and a spatial attention module, the multi-branch fusion module includes at least a two-dimensional convolution module, a compression and excitation attention module, and an h-swish activation function, and the dual-branch hybrid parallel attention module includes at least a parallel multi-head self-attention module, a convolutional block attention module, a one-dimensional layer normalization layer, and a feedforward neural network; Converting the signal to be identified into a corresponding time-frequency image; The time-frequency image is processed using a module in the neural network to obtain n interference types of interference signals in the signal to be identified.

2. The interference signal identification method based on neural network according to claim 1 is characterized in that: When the neural network includes three layers of convolutional attention modules connected in parallel, processing the time-frequency image using each module in the neural network includes: Performing a convolution operation on the time-frequency image using the two-dimensional depth convolution module to obtain a first intermediate result; Performing a channel attention calculation on the first intermediate result using the channel attention module to obtain a second intermediate result; performing a convolution operation on the second intermediate result using the two-dimensional point-by-point convolution module to obtain a third intermediate result; The spatial attention module is used to perform a spatial attention operation on the third intermediate result to obtain an output result of the convolutional attention module.

3. The interference signal identification method based on neural network according to claim 1 is characterized in that: When the neural network includes a multi-branch fusion module and other modules, the processing of the time-frequency image using the modules in the neural network includes: The output results of the three-layer convolutional attention module in parallel are concatenated into a feature map; Performing a convolution operation on the feature map using the two-dimensional convolution module to obtain a fourth intermediate result; performing an attention operation on the fourth intermediate result using the compression and stimulation attention module to obtain a fifth intermediate result; Using the h-swish activation function to operate on the fifth intermediate result to obtain a sixth intermediate result; The sixth intermediate result is processed by other modules to obtain output results of the other modules.

4. The interference signal identification method based on neural network according to claim 3 is characterized in that: When the neural network includes a dual-branch hybrid parallel attention module and a fully connected layer, processing the time-frequency image using the modules in the neural network includes: Normalizing the output results of the other modules using the one-dimensional normalization layer in the dual-branch hybrid parallel attention module, multiplying the normalized results by three different weight matrices respectively to obtain a query, a key, and a value, performing an attention operation and concatenating the query, the key, and the value using the multi-head self-attention module in the dual-branch hybrid parallel attention module to obtain a seventh intermediate result; Using the channel attention module in the convolution block attention module to perform a channel attention operation on the output results of the other modules, and using the spatial attention module in the convolution block attention module to perform a spatial attention operation on the operation results, to obtain an eighth intermediate result; performing a normalization operation on the sum of the seventh intermediate result and the eighth intermediate result using the one-dimensional layer normalization layer to obtain a ninth intermediate result; Processing the ninth intermediate result using the feedforward neural network to obtain an output result of the dual-branch hybrid parallel attention module; The fully connected layer is used to calculate the output result of the dual-branch hybrid parallel attention module to obtain n interference types of the interference signal in the signal to be identified.

5. The interference signal identification method based on neural network according to claim 1 is characterized in that: The converting the signal to be identified into a corresponding time-frequency image includes: Converting the signal to be identified into a corresponding time-frequency image based on short-time Fourier transform; or The signal to be identified is converted into a corresponding time-frequency image based on wavelet transform.

6. The interference signal identification method based on neural network according to claim 1 is characterized in that: The interference type includes at least one of single-tone interference, multi-tone interference, linear frequency sweep interference, pulse interference, comb spectrum interference, narrowband noise interference and broadband noise interference.

7. The interference signal identification method based on a neural network according to any one of claims 1 to 6, characterized in that: The method further comprises: Acquire multiple groups of training samples, each group of training samples including interference signal samples and pre-labeled labels, wherein the interference signal samples are obtained by superimposing at least a noise signal, a useful signal, and an interference signal of at least one interference type, and the labels are used to label all actual interference types corresponding to the interference signal samples; creating the neural network; For each set of training samples, using the modules in the neural network to process the interference signal samples in the training samples to obtain the corresponding predicted interference type; calculating the loss value according to the predicted interference type, the actual interference type and a preset loss function; The parameters of the neural network are adjusted according to the loss value, and the training is stopped after the parameters meet the preset conditions.

8. An interference signal recognition device based on a neural network, characterized in that: The device comprises: An acquisition module is used to acquire a signal to be identified, where the signal to be identified is obtained by superimposing at least a noise signal, a useful signal, and interference signals of n interference types, where n is an unknown number greater than or equal to 1; The acquisition module is further used to acquire a pre-trained neural network, which includes at least a three-layer convolutional attention module, a multi-branch fusion module, a dual-branch hybrid parallel attention module and a fully connected layer in parallel, the convolutional attention module includes at least a two-dimensional depth convolution module, a channel attention module, a two-dimensional point-by-point convolution module and a spatial attention module, the multi-branch fusion module includes at least a two-dimensional convolution module, a compression and excitation attention module and an h-swish activation function, and the dual-branch hybrid parallel attention module includes at least a parallel multi-head self-attention module, a convolutional block attention module, a one-dimensional layer normalization layer and a feedforward neural network; A conversion module, configured to convert the signal to be identified into a corresponding time-frequency image; An identification module is used to process the time-frequency image using a module in the neural network to obtain n interference types of interference signals in the signal to be identified.

9. A computer-readable storage medium, characterized in that The storage medium stores at least one instruction, and the at least one instruction is loaded and executed by the processor to implement the interference signal identification method based on a neural network as described in any one of claims 1 to 7.

10. A computer device, characterized in that: The computer device includes: the interference signal recognition device based on neural network as described in claim 8.

Images