Radar individual open set identification method and device based on deep learning

Abstract

The invention relates to a radar individual open set identification method and device based on deep learning, and the method comprises the steps: obtaining a to-be-identified radar individual signal; preprocessing a radar individual signal to be identified to obtain a plurality of segments of sequence signals; inputting the plurality of segments of sequence signals into a deep learning network model to obtain the category of the radar individual signal to be identified; the deep learning network model comprises a feature extraction module used for extracting feature information of a plurality of segments of sequence signals; the long-short-term memory network module is used for learning the feature relationship among the feature information of the plurality of segments of sequence signals; the multi-layer sensor module is used for performing dimensionality reduction on the features output by the long-short-term memory network module; and the classification module is used for completing identification of the radar individual signals according to the features after dimension reduction so as to obtain the category of the radar individual signals to be identified. According to the method, signals of known types can be identified and classified, and unknown electromagnetic signals can also be identified.

Description

Technical Field

[0001] The present invention relates to the field of radar signal processing technology, and in particular to a radar individual open set recognition method and device based on deep learning. Background Art

[0002] However, as the electromagnetic environment becomes increasingly complex and radar radiation sources become more diverse, the task of identifying individual radiation sources for modern electronic countermeasures is full of severe challenges. In particular, electronic reconnaissance receivers are often faced with an open electromagnetic environment, in which the types and quantities of radiation sources are constantly changing, but the signal database capacity is limited, so completely unknown new radiation sources may appear during identification. Most existing methods are basically for closed set identification, which can only identify the radar individual types already in the database, and do not have the ability of open set identification, and will mistakenly identify new radiation sources as some known radiation sources. In particular, the currently widely used deep learning-based methods, if the categories of radiation source samples not included in the training phase appear during identification, will most likely result in identification errors, seriously affecting the credibility of the identification results.

[0003] Due to the different total individual categories and the number of each individual of different radiation sources, the methods of identifying and classifying them are also different. In fact, the signals emitted by any individual have certain differences in essence, namely fingerprint features. Therefore, in theory, a deep learning network model can be designed to learn the deep features of the signal. By training a large number of samples of known types, the deep fingerprint features of this known signal can be learned. For the signals of individuals of known types, the trained model can reconstruct the features of the signal and fully extract the features of the known type of signal. The fingerprint features of the signals emitted by the same category should be highly similar, and the distance in the feature space is closer than that of other individual categories. Based on this, individuals of the same category can be identified and classified. For the signals of individuals of unknown types, the model trained with known signals cannot extract the feature fingerprints of these signals, and the performance in the feature space is significantly different from that of known types. In general, the model can shorten the feature distance of signals of known types, but it has a counter-effect on signals of unknown categories. Summary of the invention

[0004] The technical problem to be solved by the present invention is to provide a radar individual open set identification method and device based on deep learning, which can not only identify and classify signals of known types, but also identify unknown electromagnetic signals.

[0005] The technical solution adopted by the present invention to solve the technical problem is: to provide a radar individual open set identification method based on deep learning, comprising the following steps:

[0006] Acquire the radar individual signal to be identified;

[0007] Preprocessing the radar individual signal to be identified to obtain a plurality of sequence signals;

[0008] The plurality of sequence signals are input into a deep learning network model to obtain the category of the radar individual signal to be identified; wherein the deep learning network model includes:

[0009] A convolutional neural network module, used to extract feature information of the plurality of sequence signals;

[0010] Long short-term memory network module, used to learn the characteristic relationship between the characteristic information of several segments of sequence signals;

[0011] A multi-layer perceptron module, used for reducing the dimension of the features output by the long short-term memory network module;

[0012] The classification module is used to complete the identification of the radar individual signal according to the features after dimensionality reduction, and obtain the category of the radar individual signal to be identified.

[0013] The preprocessing of the radar individual signal to be identified to obtain a plurality of sequence signals specifically includes:

[0014] Converting the radar individual signal to be identified into an IQ data signal;

[0015] The IQ data signal is divided into a plurality of sequence signals.

[0016] The feature extraction module includes a first convolution layer, a first batch normalization layer, a RELU activation layer, a second convolution layer, a third convolution layer, a Dropout layer, a fourth convolution layer, a fifth convolution layer and a second batch normalization layer connected in sequence; the convolution kernels of the second convolution layer, the third convolution layer, the fourth convolution layer and the fifth convolution are all 3, and the convolution kernel of the first convolution layer is larger than the convolution kernel of the second convolution layer.

[0017] The classification module adopts a softmax classifier during the training of the deep learning network model and adopts an unsupervised clustering algorithm model during the use of the deep learning network model.

[0018] The deep learning network model is trained in the following way:

[0019] Collect different types of individual radar signals;

[0020] Preprocessing the different types of radar individual signals to obtain multiple sequence signals of each type of radar individual;

[0021] Divide multiple sequence signals into training data sets, test data sets and validation data sets;

[0022] The deep learning network model is trained using a training data set, and after the training is completed, the deep learning network model is tested using a test data set;

[0023] If the test result meets the preset requirements, the classification module in the deep learning network model is set as an unsupervised clustering algorithm model, and the parameters obtained by training are retained;

[0024] The modified deep learning network model is tested using the test data set. If the test result meets the preset requirements, the training of the deep learning network model is completed. If the test result does not meet the preset requirements, the deep learning network model continues to be trained until the test result meets the preset requirements.

[0025] The loss function of the deep learning network model is: loss = lam*CNLoss(pred,label)+(1-lam)*CNLoss(pred,label randon ), where loss is the loss function value of the deep learning network model, lam is the sample drawn from the β distribution, CNLoss() is the cross entropy loss function with the preset weight added, pred is the predicted signal category, label is the actual signal category, and label randon is the category of the signal after the order is randomly shuffled.

[0026] The technical solution adopted by the present invention to solve the technical problem is: to provide a radar individual open set identification device based on deep learning, comprising:

[0027] An acquisition module, used for acquiring the radar individual signal to be identified;

[0028] A preprocessing module, used for preprocessing the radar individual signal to be identified to obtain a plurality of sequence signals;

[0029] The identification module is used to input the plurality of sequence signals into a deep learning network model to obtain the category of the radar individual signal to be identified; wherein the deep learning network model includes:

[0030] A feature extraction module, used to extract feature information of the plurality of sequence signals;

[0031] Long short-term memory network module, used to learn the characteristic relationship between the characteristic information of several segments of sequence signals;

[0032] A multi-layer perceptron module, used for reducing the dimension of the features output by the long short-term memory network module;

[0033] The classification module is used to complete the identification of the radar individual signal according to the features after dimensionality reduction, and obtain the category of the radar individual signal to be identified.

[0034] The technical solution adopted by the present invention to solve its technical problem is: to provide an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the above-mentioned radar individual open set identification method based on deep learning when executing the computer program.

[0035] The technical solution adopted by the present invention to solve its technical problem is: providing a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the steps of the above-mentioned radar individual open set identification method based on deep learning are implemented.

[0036] Beneficial Effects

[0037] Due to the adoption of the above-mentioned technical scheme, the present invention has the following advantages and positive effects compared with the prior art: the present invention targets the signals of various radar radiation sources existing in a complex electromagnetic environment, and utilizes the designed deep learning network model to learn and classify the signals in the environment, thereby realizing effective recognition and classification of known individual types, and can also recognize radar individuals of unknown types that have never been seen before. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] Figure 1 is a flowchart of a radar individual open set identification method based on deep learning according to a first embodiment of the present invention;

[0039] Figure 2 is a schematic diagram of an IQ signal after preprocessing in the first embodiment of the present invention;

[0040] Figure 3 is a schematic diagram of a sequence signal after preprocessing in the first embodiment of the present invention;

[0041] Figure 4 Schematic diagram of the structure of the deep learning network model (closed set) in the first embodiment of the present invention;

[0042] Figure 5 is a schematic diagram of the structure of a deep learning network model (open set) in the first embodiment of the present invention;

[0043] Figure 6 is a training flow chart of a deep learning network model in the first embodiment of the present invention;

[0044] Figure 7 is a schematic diagram of the weight of the loss function in the first embodiment of the present invention;

[0045] Figure 8 This is a flow chart of accuracy calculation in the first embodiment of the present invention;

[0046] Fig. 9 This is a schematic diagram of the change of the training set loss in the first embodiment of the present invention;

[0047] Fig.10 Schematic diagram of the change of the accuracy of the closed set recognition verification set in the first embodiment of the present invention;

[0048] Fig.11 This is a partial open set recognition result diagram in the first embodiment of the present invention. DETAILED DESCRIPTION

[0049] The present invention will be further described below in conjunction with specific embodiments. It should be understood that these embodiments are only used to illustrate the present invention and are not intended to limit the scope of the present invention. In addition, it should be understood that after reading the content taught by the present invention, those skilled in the art can make various changes or modifications to the present invention, and these equivalent forms fall within the scope limited by the appended claims of the application equally.

[0050] The first embodiment of the present invention relates to a radar individual open set identification method based on deep learning, such as Figure 1 As shown, the following steps are included:

[0051] Step 1: Obtain the radar individual signal to be identified.

[0052] Step 2: pre-process the radar individual signal to be identified to obtain several sequence signals. The pre-processing in this step is specifically: using down-conversion DDC to convert the radar individual signal to be identified into an IQ data signal (see Figure 2 ); dividing the IQ data signal into a plurality of sequence signals (see Figure 3 ), the length of each segment of the sequence signal is equal when segmenting. The preprocessed IQ data signal contains more signal information, including amplitude and phase.

[0053] Step 3: input the several segments of sequence signals into a deep learning network model to obtain the category of the individual radar signal to be identified.

[0054] like Figure 4 and Figure 5As shown, the deep learning network model in this step includes: a convolutional neural network module, which is used to extract the feature information of the plurality of sequence signals; a long short-term memory network (LSTM) module, which is used to learn the feature relationship between the feature information of the plurality of sequence signals; a multi-layer perceptron (MLP) module, which is used to reduce the dimension of the features output by the long short-term memory network module; and a classification module, which is used to complete the recognition of the radar individual signal according to the features after the dimension reduction, and obtain the category of the radar individual signal to be recognized.

[0055] The feature extraction module in this embodiment uses a convolutional neural network for feature extraction, which includes a first convolutional layer, a first normalization layer, a RELU activation layer, a second convolutional layer, a third convolutional layer, a Dropout layer, a fourth convolutional layer, a fifth convolutional layer and a second normalization layer connected in sequence. In order to extract the information of the entire signal sequence, obtain greater context information, and have a larger receptive field, the first convolutional layer uses a convolutional network with a larger convolution kernel, so that the feature information of a longer sequence can be processed at one time. However, the use of a large convolutional kernel is prone to the loss of detailed features, and the convolutional network with a large convolutional kernel pays more attention to global information. Therefore, in order to allow the model to learn more refined local information, in this embodiment, the second convolutional layer, the third convolutional layer, the fourth convolutional layer and the fifth convolutional layer use a convolutional network with a small convolutional layer with a convolutional kernel of 3, and the step size of the convolutional layer is 2. At the same time, this embodiment also uses a RELU activation layer after the large convolution kernel, and the nonlinear relationship between the feature dimensions is increased by the RELU activation function. This embodiment also uses a batch normalization layer to increase the convergence speed of the model, making the model fitting more efficient. However, this also means an increased risk of overfitting the model. Therefore, a Dropout layer is used after the third convolutional layer to increase the generalization ability of the model. By randomly discarding, the model has a higher fault tolerance rate and reduces the overfitting phenomenon of the model.

[0056] For closed-set models (see Figure 4 ), the classification module uses a softmax classifier to classify the output of the MLP module, and finally calculates the error between the classification result and the actual true value, and uses the error value to optimize the model.

[0057] Open set model (see Figure 5) is obtained by further optimizing the results of the closed set model. Therefore, after the closed set model is trained, it is necessary to change the classification module of the closed set model and change the softmax classifier of the closed set model to an unsupervised clustering algorithm model to achieve the results of the open set model. In order to finally obtain the unknown type of result, the network of the trained closed set model including the parameters of the network is used to perform unsupervised clustering on the output results of the MLP module without the softmax classifier. The unsupervised clustering category is one more than the closed set model category, and the extra category is the unknown category that the open set model needs to obtain.

[0058] The deep learning network model in this embodiment is trained by the following method:

[0059] In order to realize open set recognition and classification of radar individual signals, the present invention first collects different types of radar individual signals and preprocesses the collected different types of radar individual signals, that is, converts the AD signal into an IQ data signal through down-conversion DDC, and then slices the converted IQ data signal to cut the long sequence signal into multiple small sequence signals.

[0060] After completing the above preprocessing of the data, the data set is divided into three parts, namely, a training data set, a test data set, and a validation data set. Before using the above data set to train the deep learning network model, it is also necessary to design a loss function for the deep learning network model to determine the error of target recognition classification. The loss function used in this embodiment is the cross entropy loss function (CrossEntropyLoss). It should be noted that the categories in the data set used in this embodiment are extremely unbalanced, which is also a problem that needs to be faced in actual situations. To address this problem, this embodiment sets corresponding weights for each category to balance the model's learning ability for each category. The weights are set as follows: Figure 7 When calculating the loss function, this implementation extracts samples from the β distribution as additional calculation weights:

[0061] lam=β(15,15)

[0062] Wherein, lam is a sample drawn from the β distribution. Since the number of positive and negative times set in this embodiment is the same, the probability is about 0.5. The loss function in this embodiment is expressed as:

[0063] loss=lam*CNLoss(pred,label)+(1-lam)*CNLoss(pred,label) randon )

[0064] CNLoss() is added Figure 7The weighted cross entropy loss function is: pred is the predicted signal category, label is the actual signal category, label randon is the category of the signal after the order is randomly shuffled.

[0065] After setting the loss function, the data is trained based on the deep learning network model. The Adam optimizer is used, the learning rate of the model is set to 0.0001, and the batch size is 8. The model training process of this implementation is as follows: Figure 6 As shown, specifically: the deep learning network model is trained using a training data set, and after the training is completed, the deep learning network model is tested using a test data set; if the test result meets the preset requirements, the classification module in the deep learning network model is set to an unsupervised clustering algorithm model, and the parameters obtained by training are retained; the changed deep learning network model is tested using the test data set, and if the test result meets the preset requirements, the training of the deep learning network model is completed, and if the test result does not meet the preset requirements, the deep learning network model continues to be trained until the test result meets the preset requirements.

[0066] After the samples are labeled and classified, they are sent to the deep learning network model for training and recognition, and the following can be obtained: Fig. 9 and Fig.10 After the network is trained, new samples are identified and classified online (including samples of individuals of unknown individual categories) to verify the performance of the model. New signal samples are sent to the deep learning model for detection, and the following is obtained: Fig.11 Results shown.

[0067] It can be seen that after learning the known samples in the signal sequence library, the deep learning network model of this embodiment will automatically find the fingerprint features of the signal data of known categories, shorten the feature distance between the fingerprint features of the same category, and the model cannot effectively learn the fingerprint features of unknown categories, so the feature distance in the feature space is farther. Through the recognition and classification of the features, a model network with the ability to recognize and classify individual radar signals is formed, and the network can be used to recognize and classify individual radar signals.

[0068] The performance parameters of the deep learning network model of this implementation are evaluated below, and the accuracy of the model is calculated as follows:

[0069] TP+TN

[0070] accuracy=TP+TN+FP+FN

[0071] Among them, TP is the retrieved positive example, TN is the unretrieved negative example, FN is the unretrieved positive example, and FP is the retrieved negative example.

[0072] Since this embodiment divides a signal into multiple small sample signals during data preprocessing, but the accuracy is calculated based on the entire large signal segment, a voting method is adopted to determine which category the long sequence signal belongs to based on which category of small samples is the largest in a large signal segment. The process is as follows: Figure 8 Then, the accuracy of the validation set and the test set is calculated according to the above formula to test the recognition effect of the model.

[0073] It is not difficult to find that the present invention targets the signals of various radar radiation sources in a complex electromagnetic environment, and utilizes the designed deep learning network model to learn and classify the signals in the environment, thereby achieving effective recognition and classification of known individual types, and can also identify radar individuals of unknown types that have never been seen before.

[0074] A second embodiment of the present invention relates to a radar individual open set identification device based on deep learning, comprising:

[0075] An acquisition module, used for acquiring the radar individual signal to be identified;

[0076] A preprocessing module, used for preprocessing the radar individual signal to be identified to obtain a plurality of sequence signals;

[0077] The identification module is used to input the plurality of sequence signals into a deep learning network model to obtain the category of the radar individual signal to be identified; wherein the deep learning network model includes:

[0078] A feature extraction module, used to extract feature information of the plurality of sequence signals;

[0079] Long short-term memory network module, used to learn the characteristic relationship between the characteristic information of several segments of sequence signals;

[0080] A multi-layer perceptron module, used for reducing the dimension of the features output by the long short-term memory network module;

[0081] The classification module is used to complete the identification of the radar individual signal according to the features after dimensionality reduction, and obtain the category of the radar individual signal to be identified.

[0082] The preprocessing module comprises:

[0083] A conversion unit, used for converting the radar individual signal to be identified into an IQ data signal;

[0084] The slicing unit is used to slice the IQ data signal into a plurality of sequence signals.

[0085] The feature extraction module includes a first convolution layer, a first normalization layer, a RELU activation layer, a second convolution layer, a third convolution layer, a Dropout layer, a fourth convolution layer, a fifth convolution layer and a second normalization layer connected in sequence; the convolution kernels of the second convolution layer, the third convolution layer, the fourth convolution layer and the fifth convolution are all 3, and the convolution kernel of the first convolution layer is larger than the convolution kernel of the second convolution layer.

[0086] The classification module adopts a softmax classifier during the training of the deep learning network model and adopts an unsupervised clustering algorithm model during the use of the deep learning network model.

[0087] The deep learning network model is trained in the following way:

[0088] Collect different types of individual radar signals;

[0089] Preprocessing the different types of radar individual signals to obtain multiple sequence signals of each type of radar individual;

[0090] Divide multiple sequence signals into training data sets, test data sets and validation data sets;

[0091] The deep learning network model is trained using a training data set, and after the training is completed, the deep learning network model is tested using a test data set;

[0092] If the test result meets the preset requirements, the classification module in the deep learning network model is set as an unsupervised clustering algorithm model, and the parameters obtained by training are retained;

[0093] The modified deep learning network model is tested using the test data set. If the test result meets the preset requirements, the training of the deep learning network model is completed. If the test result does not meet the preset requirements, the deep learning network model continues to be trained until the test result meets the preset requirements.

[0094] The loss function of the deep learning network model is: loss = lam*CNLoss(pred,label)+(1-lam)*CNLoss(pred,label randon ), where loss is the loss function value of the deep learning network model, lam is the sample drawn from the β distribution, CNLoss() is the cross entropy loss function with the preset weight added, pred is the predicted signal category, label is the actual signal category, and label randon is the category of the signal after the order is randomly shuffled.

[0095] A third embodiment of the present invention relates to an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the steps of the deep learning-based radar individual open set identification method of the first embodiment are implemented.

[0096] A fourth embodiment of the present invention relates to a computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, the steps of the radar individual open set identification method based on deep learning of the first embodiment are implemented.

[0097] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage and optical storage, etc.) that contain computer-usable program code.

[0098] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0099] These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer-readable memory produce a manufactured product including an instruction method, which is implemented in the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0100] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0101] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

Claims

1. A radar individual open set identification method based on deep learning, characterized in that: The following steps are involved: Acquire the radar individual signal to be identified; Preprocessing the radar individual signal to be identified to obtain a plurality of sequence signals; The plurality of sequence signals are input into a deep learning network model to obtain the category of the radar individual signal to be identified; wherein the deep learning network model includes: A feature extraction module, used to extract feature information of the plurality of sequence signals; Long short-term memory network module, used to learn the characteristic relationship between the characteristic information of several segments of sequence signals; A multi-layer perceptron module, used for reducing the dimension of the features output by the long short-term memory network module; The classification module is used to complete the identification of the radar individual signal according to the features after dimensionality reduction, and obtain the category of the radar individual signal to be identified.

2. The radar individual open set identification method based on deep learning according to claim 1 is characterized in that: The preprocessing of the radar individual signal to be identified to obtain a plurality of sequence signals specifically includes: Converting the radar individual signal to be identified into an IQ data signal; The IQ data signal is divided into a plurality of sequence signals.

3. The radar individual open set identification method based on deep learning according to claim 1 is characterized in that: The feature extraction module includes a first convolution layer, a first batch normalization layer, a RELU activation layer, a second convolution layer, a third convolution layer, a Dropout layer, a fourth convolution layer, a fifth convolution layer and a second batch normalization layer connected in sequence; the convolution kernels of the second convolution layer, the third convolution layer, the fourth convolution layer and the fifth convolution are all 3, and the convolution kernel of the first convolution layer is larger than the convolution kernel of the second convolution layer.

4. The radar individual open set identification method based on deep learning according to claim 1 is characterized in that: The classification module adopts a softmax classifier during the training of the deep learning network model and adopts an unsupervised clustering algorithm model during the use of the deep learning network model.

5. The radar individual open set identification method based on deep learning according to claim 4 is characterized in that: The deep learning network model is trained in the following way: Collect different types of individual radar signals; Preprocessing the different types of radar individual signals to obtain multiple sequence signals of each type of radar individual; Divide multiple sequence signals into training data sets, test data sets and validation data sets; The deep learning network model is trained using a training data set, and after the training is completed, the deep learning network model is tested using a test data set; If the test result meets the preset requirements, the classification module in the deep learning network model is set as an unsupervised clustering algorithm model, and the parameters obtained by training are retained; The modified deep learning network model is tested using the test data set. If the test result meets the preset requirements, the training of the deep learning network model is completed. If the test result does not meet the preset requirements, the deep learning network model continues to be trained until the test result meets the preset requirements.

6. The radar individual open set identification method based on deep learning according to claim 1 is characterized in that: The loss function of the deep learning network model is: loss = lam*CNLoss(pred,label)+(1-lam)*CNLoss(pred,label randon ), where loss is the loss function value of the deep learning network model, lam is the sample drawn from the β distribution, CNLoss() is the cross entropy loss function with the preset weight added, pred is the predicted signal category, label is the actual signal category, and label randon is the category of the signal after the order is randomly shuffled.

7. A radar individual open set identification device based on deep learning, characterized in that: include: An acquisition module, used for acquiring the radar individual signal to be identified; A preprocessing module, used for preprocessing the radar individual signal to be identified to obtain a plurality of sequence signals; The identification module is used to input the plurality of sequence signals into a deep learning network model to obtain the category of the radar individual signal to be identified; wherein the deep learning network model includes: A feature extraction module, used to extract feature information of the plurality of sequence signals; Long short-term memory network module, used to learn the characteristic relationship between the characteristic information of several segments of sequence signals; A multi-layer perceptron module, used for reducing the dimension of the features output by the long short-term memory network module; The classification module is used to complete the identification of the radar individual signal according to the features after dimensionality reduction, and obtain the category of the radar individual signal to be identified.

8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the steps of the radar individual open set identification method based on deep learning as described in any one of claims 1 to 6 are implemented.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the radar individual open set identification method based on deep learning as described in any one of claims 1 to 6 are implemented.

Images