Radar PRI modulation type identification method based on neural network and multi-head attention mechanism

Abstract

The invention relates to a radar PRI modulation type identification method based on a neural network and a multi-head attention mechanism, and belongs to the field of radar signal processing. First-order difference is carried out on a radar pulse TOA sequence to obtain sample data of a PRI sequence, a median filter is adopted to carry out filtering processing on the sample data to reduce the influence of missing or false pulses, and a model integrating a multi-scale convolution extrusion excitation neural network, a bidirectional gating circulation unit neural network and a multi-head attention mechanism is constructed. The model can extract data local refinement features and capture data long-term time sequence correlation features, a multi-head attention mechanism is used for integrating global information, more weights are applied to important time steps in a gating circulation unit, the model is trained and tested, and the result shows that compared with other neural network methods, the model has the advantages that the algorithm is simple, and the algorithm is easy to implement. According to the method, the recognition rate of six kinds of radar PRI modulation signals is higher under the condition that the ratio of missing pulses to stray pulses is high.

Description

Technical Field

[0001] The present invention relates to the field of radar signal processing, and in particular to classification and recognition of pulse-to-pulse modulation of radar signals, and in particular to a method for identifying PRI modulation types of radar signals based on deep learning. Background Art

[0002] In modern electronic warfare, the complexity and diversification of radar technology make the pulse repetition interval (PRI) modulation pattern recognition of radar signals particularly important. Radar warning receivers (RWR) and electronic support (ESM) systems rely on the recognition of PRI modulation types to identify the emitter, which is crucial for model identification, type determination, working mode identification, and threat assessment. By accurately identifying the PRI modulation pattern, not only can we build a database for radiator identification, but also provide an important basis for electromagnetic environment situation awareness and the formulation of interference strategies. With the development of new radar systems, the PRI modulation pattern of radar signals has become increasingly complex. In addition to the traditional fixed pattern, it also includes a variety of complex modulation forms such as staggered, jitter, group change, sliding and sine modulation. These complex PRI modulation patterns have brought huge challenges to radar signal recognition technology, and new recognition methods and technologies need to be continuously developed to cope with them. Effective PRI modulation pattern recognition technology can improve the detection and recognition capabilities of electronic warfare systems, thereby taking an active position in the modern battlefield.

[0003] Generally speaking, the recognition methods of radar signal PRI modulation mode can be divided into traditional methods based on conventional parameters and methods based on deep learning. Traditional recognition methods are mainly histogram method, PRI transformation method and their variants. These methods can realize the classification and recognition of simple radar systems by counting the number of pulse distribution or combining other manually extracted features. The method based on deep learning can more effectively process high-dimensional and complex structure data and automatically extract rich feature expressions. It has stronger generalization ability, can learn the potential distribution of data, and has scalability and flexibility. By splicing different types of neural network algorithms, better recognition effect can be achieved. At present, domestic and foreign scholars have proposed some radar PRI modulation mode recognition methods based on deep learning, such as methods based on long short-term memory (LSTM) network, methods based on convolutional neural network (CNN), methods based on time domain convolutional neural network (TCN), etc. However, the recognition accuracy of these methods still needs to be improved when the pulse missing rate and false pulse ratio are high. Summary of the invention

[0004] The present invention provides a radar PRI modulation type recognition method based on a neural network and a multi-head attention mechanism, so as to solve the problems of relying on experience-based threshold selection and low recognition accuracy existing in the existing radar PRI modulation type recognition method.

[0005] The technical solution adopted by the present invention comprises the following steps:

[0006] Step 1, simulate and generate six types of pulse-interval modulation radar signals, including fixed modulation signal, staggered modulation signal, jitter modulation signal, group-variable modulation signal, sinusoidal modulation signal and sliding modulation signal, perform noise processing on the above six types of pulse-interval modulation radar signals with different proportions of missing pulses and false pulses, and obtain the radar signal arrival time TOA sequence data set;

[0007] Step 2: Perform a differential calculation on the radar signal TOA sequence to obtain the pulse repetition interval PRI sequence of the radar signal. Then, use a median filter to filter the noisy PRI sequence containing false pulses and missing pulses, and then divide all PRI sequences into a training set and a test set.

[0008] Step 3, design a multi-scale convolution squeeze excitation neural network MSCSE module to replace the general convolution neural network module, and fuse the extracted local features with different scales and different weights into a feature vector to input into the next layer of neural network;

[0009] Step 4: Based on step 3, a radar PRI modulation recognition algorithm model based on a neural network and a multi-head attention mechanism is constructed. The model includes an MSCSE module, a bidirectional gated unit BiGRU network module, and a multi-head attention mechanism MHA module. The BiGRU network simultaneously utilizes the context features of the forward and backward directions in the sequence. The MHA module highlights the key information of the BiGRU network at different time steps, enhancing the algorithm model's ability to focus on important information in the data.

[0010] Step 5: Send the generated PRI sequence training set into the algorithm model for model training, adjust the model's hyperparameters, update the model's weights and biases, and use the trained algorithm model to send the test sets under different noise conditions into the model for classification and recognition, so as to obtain the recognition rate of the radar signal PRI modulation mode under different noise conditions.

[0011] In step 1 of the present invention, six PRI modulated radar signals are simulated and generated, and the time interval between the arrival of adjacent pulses is the pulse repetition interval PRI, which is defined by a function as follows:

[0012] P={p1,p2,…,p N},p i =t i+1 -t i

[0013] Where P is the PRI sequence, {t i,i=1,2,…,N+1} is the arrival time of each pulse signal in the radar pulse train being analyzed, i is the data sequence number, N is the number of radar pulse repetition intervals contained in the radar pulse train, and there are many different working modes and modulation modes for PRI in different radar systems. In practical applications, due to the influence of noise, the PRI sequence has a high loss and spurious pulse rate. Random noise is added to the TOA sequence of the generated six PRI modulation signals, including pulse missing noise and pulse false noise. The missing pulse sequence can be expressed as:

[0014]

[0015] Among them, {p′ n} indicates the received TOA sequence T n The new PRI sequence when j pulses are missing from the i-th pulse to the i+j-1 pulse, the false pulse sequence can be expressed as:

[0016]

[0017] {p′ n ′} represents a new PRI sequence with j pulses added after the i-th pulse, T i is the interval between the last false pulse and the next true pulse of the ith pulse.

[0018] The PRI of different radar systems described in the present invention has a variety of different working modes and modulation modes, including the following 6 types of PRI modulation type signals:

[0019] Fixed PRI modulation signals generally have a fixed value, and their fluctuation range does not exceed 1%;

[0020] In the staggered PRI modulated signal, multiple PRI values ​​appear in a periodic sequence;

[0021] In a jittered PRI modulated signal, the value of the PRI sequence jitters around a constant value. The jitter range is usually randomly selected between 3% and 30%, usually following a Gaussian distribution.

[0022] The value of the slip modulation PRI sequence usually increases monotonically from an initial value until it reaches a maximum value, or decreases monotonically from a maximum value to a minimum value;

[0023] The numerical variation law of the sinusoidal modulation PRI sequence has a shape similar to that of a sine function and appears periodically;

[0024] Group-variable modulation is similar to staggered modulation, with several fixed values. The difference is that different PRI values ​​in group-variable modulation have a certain residence time, and the number of PRI residence arrays also changes periodically.

[0025] In step 2 of the present invention, {p′ n},{p′ n The first-order difference calculation is performed on each inter-pulse modulation TOA sequence of ′} to obtain the PRI sequence, which is first normalized to avoid a large difference in the order of magnitude of different data and to speed up the calculation. The normalization process is expressed as:

[0026]

[0027] Among them, s max is the maximum value in the current sequence, s min is the minimum value in the current sequence. Then, each normalized PRI sequence is subjected to median filtering. The median filter can effectively suppress the noise caused by missing pulses and false pulses. For a window of size 2K+1, it is expressed as:

[0028] x i =median{y i-K ,y i-K+1 ,…,y i+K}

[0029] where x i It is the data point after filtering. The new PRI sequence data after median filter processing is divided into training set and test set. The ratio of training set to data set is 8:2.

[0030] The total number of samples in the dataset of the present invention is 60,000.

[0031] In step 3 of the present invention, a multi-scale convolution module MSCSE integrating the squeeze-excited attention mechanism is designed. First, four convolution kernels of different sizes, 1×1, 1×3, 1×5, and 1×7, are used to process data. The model can capture features of different scales. Smaller convolution kernels focus on local fine-grained features, while larger convolution kernels can capture longer-range dependencies, avoiding information loss that may be caused by relying only on a single-scale convolution kernel. The calculation formula for a convolution kernel size of 1×k can be expressed as:

[0032]

[0033] Among them, a represents the ath channel under the convolution kernel operation, Represents the convolution weight of the ath channel of the convolution kernel, Indicates the corresponding offset, X t is a subsequence of the input data intercepted at time step t, is the corresponding output result. After the convolution process, the output result Y is obtained. kAfter batch normalization, the output of each convolutional layer is concatenated to obtain the total output Y of a multi-scale convolution. The concatenation operation formula is as follows:

[0034] Y=Concat(Y1,Y3,Y5,Y7)

[0035] Subsequently, through the squeezing and excitation process, the module can adaptively adjust the weights of the feature channels so that the network pays more attention to important features. First, in the squeezing process, the input feature map dimension is H*1*C, which represents the length, width and number of channels of the feature map respectively. Global average pooling is performed on each channel to compress the feature map of each channel into a value, namely 1*1*C, to obtain a global description of each channel. The operation expression is as follows:

[0036]

[0037] Among them, F sq Represents the squeeze operation, Y(i) represents the input feature of the i-th channel, followed by the excitation operation, which inputs the channel description vector into a small fully connected network, including two fully connected layers FC and an activation function Sigmoid, to generate the weight coefficient of each channel. The weight coefficient is the scale factor for adjusting the original feature map. The calculation process expression is as follows:

[0038] ω c =F ex (Z c ,W)=Sigmoid(W2(W1Z c +b1)+b2)

[0039] Among them, F ex represents the excitation operation, W1 and b1 are the weight and bias of the first fully connected layer, W2 and b2 are the weight and bias of the second fully connected layer, and finally, the learned weight coefficient is weighted to each channel feature, and the result is added to the input feature to get the final output. The calculation expression is as follows:

[0040] Y'=ω c Y+Y

[0041] Afterwards, the result is processed with the P-ReLU activation function. P-ReLU is an improved activation function that introduces learnable parameters based on the ReLU activation function to avoid the inability to learn effective features in the negative value area to enhance the expressiveness of the model. The calculation expression is as follows:

[0042]

[0043] X'=f(Y')

[0044] Among them, α is a learnable parameter, and then the result X' processed by the P-ReLU activation function is sent to the next layer of the network.

[0045] In step 4 of the present invention, a radar PRI modulation recognition algorithm model is constructed. The preprocessed PRI sequence data first passes through two MSCSE layers to extract features at different time scales, and then passes through a pooling layer to downsample the features to reduce the data dimension while retaining important information to further improve the calculation efficiency. Then, after passing through two BiGRU layers, the model can better learn the temporal relationship in the sequence and capture the context information. Subsequently, after passing through the multi-head attention layer model, the model can focus on the most important part of the PRI sequence, making the information transmission more efficient. The data is flattened through the Flatten layer. Finally, the extracted features are integrated through a fully connected layer, and the classification results are finally output through the Softmax layer, thereby realizing accurate recognition of different types of PRI modulation.

[0046] In step 5 of the present invention, the generated PRI sequence training set is sent to the algorithm model for model training. The training process includes setting hyperparameters and updating network weights. Assuming that the final prediction result of the model is p q , the loss function used is the root mean square error function, and the calculation expression is as follows:

[0047]

[0048] Where B is the total number of batch samples, which is the same as the value of the hyperparameter Batch_size, and l q is the label value of the qth sample, p q is the predicted value of the qth sample, and RMSE is the loss function value, which is used for back-propagation to update the model weights.

[0049] The model parameters of the present invention are set as follows: the optimizer uses Adam, the batch size sample batch number is 128, the epochs iteration number is 50, and the Lr learning rate is 0.001. Afterwards, the trained prediction model is tested using the test set to verify that the model has a high prediction accuracy.

[0050] The advantage of the present invention is to provide a new algorithm model for identifying the PRI modulation mode of radar signals. The algorithm model mainly includes a multi-scale convolution squeeze excitation neural network MSCSE module, a bidirectional gated unit BiGRU network module and a multi-head attention mechanism MHA module, and is compared with the existing CNN-LSTM, TCN, and CNN network structures. Under the conditions of different missing pulses and false pulse proportions, different pulse-to-pulse modulation signals are tested to verify the effectiveness of the method. The method integrates the multi-scale convolution squeeze excitation neural network MSCSE, the bidirectional gated recurrent unit neural network and the multi-head attention mechanism, and can extract local refined features of data and capture long-term temporal correlation features of data. The global information is integrated through the multi-head attention mechanism, and more weights can be applied to important time steps in the gated recurrent unit. The classification and recognition effect of the radar signal PRI modulation mode is better under the condition that the missing pulse and false pulse proportion are high, and the accurate recognition of the radar signal PRI modulation type in a noisy environment can be achieved. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] Figure 1 It is a flow chart of the radar signal PRI modulation mode identification method in the present invention;

[0052] Figure 2 It is a schematic diagram of the PRI sequence of the fixed modulation signal modulation mode;

[0053] Figure 3 It is a schematic diagram of the PRI sequence of the jitter modulation signal modulation method;

[0054] Figure 4 It is a schematic diagram of the PRI sequence of the staggered modulation signal modulation method;

[0055] Figure 5 It is a schematic diagram of the PRI sequence of the group variable modulation signal modulation method;

[0056] Figure 6 It is a schematic diagram of the PRI sequence of the sinusoidal modulation signal modulation method;

[0057] Figure 7 It is a schematic diagram of the PRI sequence of the sliding modulation signal modulation method;

[0058] Figure 8 It is a schematic diagram of the module structure of the proposed multi-scale convolutional neural network fused with the squeeze incentive mechanism;

[0059] Fig. 9 It is the overall structural diagram of the proposed radar PRI modulation recognition algorithm model;

[0060] Fig.10 is a schematic diagram of the gated recurrent unit structure;

[0061] Fig.11 is a schematic diagram of the bidirectional gated recurrent unit structure;

[0062] Fig.12 This is a schematic diagram of the structure of the multi-head attention mechanism for hidden state features;

[0063] Fig.13 It is the recognition rate of six radar PRI modulation signals obtained by simulation when the missing pulse accounts for 0-50% and the false pulse accounts for 0-40%;

[0064] Fig.14 This is a comparison chart of the recognition rates of CNN-LSTM, TCN, CNN and the method of the present invention under different noise conditions. DETAILED DESCRIPTION

[0065] The technical solution of the present invention is described in detail below in conjunction with specific implementation methods. Figure 1 As shown, the present invention provides a radar signal PRI modulation type recognition method based on a neural network and a multi-head attention mechanism, and the specific steps are as follows:

[0066] Step 1: Simulate and generate six pulse-to-pulse modulation radar signals. The time interval between the arrival of adjacent pulses is the pulse repetition interval PRI, which is defined by the function as follows:

[0067] P={p1,p2,…,p N},p i =t i+1 -t i

[0068] Where P is the PRI sequence, {t i , i=1,2,…,N+1} is the arrival time of each pulse signal in the radar pulse train being analyzed, i is the data sequence number, and N is the number of radar pulse repetition intervals contained in the radar pulse train. The PRI of different radar systems may have a variety of different operating modes and modulation modes. The following introduces the six types of PRI modulation type signals generated:

[0069] (1) Fixed PRI modulation type, see Figure 2 , the fixed PRI modulation signal has a fixed value, and its fluctuation range will not exceed 1%. Its signal law can be expressed as:

[0070] p i =a

[0071] Among them, a is a constant, and the fixed PRI modulation type is generally used in conventional search and tracking radar tasks;

[0072] (2) Different PRI modulation types, see Figure 3In the staggered PRI modulation signal, multiple PRI values ​​appear in a periodic sequence, and the signal rule can be expressed as:

[0073] p i =p u ,i=umodM

[0074] Wherein, M refers to the number of PRIs in one cycle, u = 1, 2, ..., M, each cycle is called a skeleton cycle, and the staggered PRI signal is often used to eliminate blind speed in MTI radar (moving target detection radar);

[0075] (3) Jitter PRI modulation type, see Figure 4 , the value of the jitter-modulated PRI sequence jitters around a certain constant value. The jitter range is usually randomly selected between 3% and 30%, usually following a Gaussian distribution. Its signal law can be expressed as:

[0076] p i =p0*(1+δ)

[0077] Among them, δ is the jitter amplitude percentage, and its value range is 3%≤|δ|≤30%. The jittered PRI modulation signal is generally used to counter interference and anti-reconnaissance;

[0078] (4) Sliding PRI modulation type, see Figure 5 , the value of the slip modulation PRI sequence increases monotonically from an initial value until it reaches the maximum value, or decreases monotonically from the maximum value to the minimum value. After reaching the maximum or minimum value, the PRI will immediately reset to the initial value, and then repeat the process. Its monotonically increasing signal law can be expressed as:

[0079]

[0080] Among them, v represents the rate of increase, l represents the length of a change cycle, and p min It is the minimum value within the cycle. The PRI variation range of sliding modulation is generally 10% to 30% above and below the center value. The purpose of sliding PRI modulation is generally to eliminate blind distance;

[0081] (5) Sinusoidal PRI modulation type, see Figure 6 , the numerical variation law of the sinusoidal modulation PRI sequence has a shape similar to that of a sine function and appears periodically. Its signal law can be expressed as:

[0082]

[0083] Where A is the amplitude of the modulation, θ is the frequency of the sine function, is the phase of the sine function, o is the number of data in a sine cycle, λ oIt is a value proportional to o, defining the sampling resolution. Sine PRI modulated signals are generally used for missile guidance, avoiding obscuration and ranging;

[0084] (6) Group variable PRI modulation type, see Figure 7 , Group-variable modulation is similar to staggered modulation, with several fixed values. The difference is that different PRI values ​​in group-variable modulation have a certain dwell time, and several PRI dwell arrays also have periodic changes. The signal law can be expressed as:

[0085] PRI(N)=PRI0+N*△PRI

[0086] Among them, PRI(N) represents the Nth group of PRI values, PRI0 represents the first group of PRI values, N represents the number of groups (counting from 0), △PRI represents the increment of the pulse repetition interval between each group, and group-variable PRI modulation can analyze the problem of distance and speed ambiguity;

[0087] In practical applications, due to the influence of noise, the PRI sequence has a high loss and spurious pulse rate. Random noise is added to the TOA sequence of the six generated PRI modulation signals, including pulse missing noise and pulse false noise. The missing pulse sequence can be expressed as:

[0088]

[0089] Among them, {p′ n} indicates the received TOA sequence T n The new PRI sequence when j pulses are missing from the i-th pulse to the i+j-1 pulse, the false pulse sequence can be expressed as:

[0090]

[0091] {p′ n ′} represents a new PRI sequence with j pulses added after the i-th pulse, T i is the interval between the last false pulse and the next true pulse of the ith pulse;

[0092] Step 2: include {p′ n},{p′ n The first-order difference calculation is performed on each inter-pulse modulation TOA sequence of ′} to obtain the PRI sequence, which is first normalized to avoid a large difference in the order of magnitude of different data and to speed up the calculation. The normalization process is expressed as:

[0093]

[0094] Among them, s max is the maximum value in the current sequence, s minis the minimum value in the current sequence, s i is the i-th processed data point. Then, each normalized PRI sequence is subjected to median filtering. The median filter can effectively suppress the noise caused by missing pulses and false pulses. For a window of size 2K+1, it is expressed as:

[0095] x i =median{y i-K ,y i-K+1 ,…,y i+K}

[0096] where x i It is the data point after filtering. The new PRI sequence data after median filter processing is divided into training set and test set. The total number of samples in the data set is 60,000, and the ratio of training set to data set is 8:2.

[0097] Step 3: Design a multi-scale convolutional module that integrates the squeeze-incentive attention mechanism. Its structure is as follows: Figure 8 As shown. First, four different sizes of convolution kernels, 1×1, 1×3, 1×5, and 1×7, are used to process the data. The model can capture features of different scales. Smaller convolution kernels focus on local fine-grained features, while larger convolution kernels can capture longer-range dependencies, avoiding information loss that may be caused by relying solely on a single-scale convolution kernel. The calculation formula for a convolution kernel size of 1×k can be expressed as:

[0098]

[0099] Among them, a represents the ath channel under the convolution kernel operation, Represents the convolution weight of the ath channel of the convolution kernel, Indicates the corresponding offset, X t is a subsequence of the input data intercepted at time step t, is the corresponding output result. After the convolution process, the output result Y is obtained. k After batch normalization and maximum pooling, the output of each convolution layer is concatenated to obtain a total output Y of a multi-scale convolution. The concatenation operation formula is as follows:

[0100] Y=Concat(Y1,Y3,Y5,Y7)

[0101] Subsequently, through the squeezing and excitation process, the module can adaptively adjust the weights of the feature channels so that the network pays more attention to important features. First, in the squeezing process, the input feature map dimension is H*1*C, which represents the length, width and number of channels of the feature map respectively. Global average pooling is performed on each channel to compress the feature map of each channel into a value, namely 1*1*C, to obtain a global description of each channel. The operation expression is as follows:

[0102]

[0103] Among them, F sq Represents the squeeze operation, Y(i) represents the input feature of the i-th channel, followed by the excitation operation, which inputs the channel description vector into a small fully connected network, including two fully connected layers FC and an activation function Sigmoid, to generate the weight coefficient of each channel. The weight coefficient is the scale factor for adjusting the original feature map. The calculation process expression is as follows:

[0104] ω c =F ex (Z c ,W)=Sigmoid(W2(W1Z c +b1)+b2)

[0105] Among them, F ex represents the excitation operation, W1 and b1 are the weight and bias of the first fully connected layer, W2 and b2 are the weight and bias of the second fully connected layer, and finally, the learned weight coefficient is weighted to each channel feature, and the result is added to the input feature to get the final output. The calculation expression is as follows:

[0106] Y'=ω c Y+Y

[0107] Afterwards, the result is processed with the P-ReLU (Parametric ReLU) activation function. P-ReLU is an improved activation function that introduces learnable parameters based on the ReLU activation function to avoid the inability to learn effective features in the negative value area to enhance the expressiveness of the model. The calculation expression is as follows:

[0108]

[0109] X'=f(Y')

[0110] Among them, α is a learnable parameter, and then the result X' processed by the P-ReLU activation function is sent to the next layer of the network;

[0111] Step 4: Based on step 3, a radar PRI modulation recognition algorithm model based on neural network and multi-head attention mechanism is constructed. The overall structure is as follows: Fig. 9 As shown in the figure. The preprocessed PRI sequence data first passes through two MSCSE layers to extract features at different time scales, then passes through a pooling layer to downsample the features, reduce the data dimension, and retain important information to further improve the computational efficiency. Then, it passes through two BiGRU layers, so that the model can better learn the temporal relationship in the sequence and capture the context information. Then, it passes through the multi-head attention layer model to focus on the most important part of the PRI sequence, making the information transmission more efficient. Finally, it passes through a fully connected layer to integrate the extracted features, and finally outputs the classification results through the Softmax layer, thereby realizing accurate recognition of different types of PRI modulation.

[0112] The main advantage of GRU is that it solves the long dependency problem in traditional RNN networks through the gating mechanism. Compared with LSTM, GRU has a simpler structure and only contains update gates and reset gates. Fig.10 As shown in Figure 2. The update gate is responsible for controlling the information transfer between the hidden state at the current moment and the hidden state at the previous moment, which helps capture long-term dependencies in the sequence. The reset gate controls how to combine the hidden state at the previous moment with the current input. When the reset gate is close to 0, GRU tends to ignore the previous state, allowing the network to pay more attention to the input information at the current moment, which helps capture short-term dependencies. Suppose the input information of GRU at time step t is X t ', reset the gate to R t , update the gate to Z t , the current hidden state is H t , the candidate hidden state is Update the calculation expression as follows:

[0113] R t =Sigmoid(W xr X t '+W hr H t-1 +b r )

[0114] Z t =Sigmoid(W xz X t '+W hz H t-1 +b z )

[0115]

[0116] Among them, W xr , W hr , W xz, W hz , W xh and W hh is the weight update matrix, b r , b z and b h is the bias, tanh is the activation function, and W is the Hadamard product operator.

[0117] BiGRU Fig.11 As shown in the figure, it consists of two GRUs with opposite directions, one responsible for forward propagation of information and the other responsible for backward propagation of information. It can keep the association between future information and past information at the same time, so it performs well in prediction and recognition. It can capture dependencies at different time scales and provide more contextual information.

[0118] The specific calculation expression is:

[0119]

[0120] in, is the hidden state at time step t in the positive direction, is the hidden state at the positive direction t-1 time step, is the hidden state at t time steps in the reverse direction, is the hidden state of the t-1 time step in the reverse direction, H t is the merged result of the hidden states of the two directions at t time steps. The merged result H of the hidden state output of each time step of BiGRU is used as the input of the next layer of multi-head attention mechanism. The attention mechanism is used to further extract contextual information related to the target task and enhance the model's ability to focus on important information. Through multi-head parallel computing, the model can focus on different parts of the input sequence from multiple different angles, such as Fig.12 As shown, for the rth attention head, the query vector (Query) Q is first calculated r 、Key vector (Key) K r Sum value vector (Value) V r The linear transformation of in, is the three vector linear transformation matrix for the rth head, then the attention output of the ith head is calculated as

[0121]

[0122] where d K is the key vector dimension. Then concatenate the outputs of all attention heads and obtain the final attention output after a linear transformation. The calculation expression is:

[0123] Multihead(Q,K,V)=Concat(head1,head2,...,head h )W O

[0124] where h is the number of attention heads, set to 8, and W O is the weight matrix of the multi-head output. Then, the Flatten layer is used to flatten the output of the multi-head attention layer, and then the fully connected layer and the Softmax function are used to obtain the final prediction result of the model, that is, the radar PRI modulation type category label.

[0125] Step 5: Send the generated PRI sequence training set to the algorithm model for model training. The training process includes setting hyperparameters and updating network weights. Assume that the final prediction result of the model is p q , the loss function used in the present invention is the root mean square error function, and the calculation expression is as follows:

[0126]

[0127] Where B is the total number of batch samples, which is the same as the value of the hyperparameter Batch_size, and l q is the label value of the qth sample, p q is the predicted value of the qth sample, RMSE is the loss function value, which is used for back-propagation to update the model weights. The model parameters are set as follows: the optimizer uses Adam, the batch size is 128, the number of Epochs is 50, and the Lr learning rate is 0.001. Afterwards, the trained prediction model is tested using the test set to verify that the model has a high prediction accuracy.

[0128] The beneficial effects of the present invention are further described below in combination with simulation experiments and results. The simulation tools for all simulation experiments are Matlab and Python 3.8.

[0129] First, the modulation parameters and added noise of the six PRI sequence data sets are introduced: the PRI value range is 100-3000us, the change rate of the sliding modulation is 20-100us, the number of different PRIs of the staggered modulation ranges from 3 to 10, the number of PRI switches of the group modulation ranges from 3 to 10, the jitter range of the jitter modulation ranges from 3% to 30%, and the amplitude change range of the sine modulation ranges from 10% to 20%. The pulse loss ratio is set to 0-50%, and the simulation step is 10%; the false pulse ratio is set to 0-40%, and the simulation step is 8%. There are 0 to 5 non-ideal reception scenarios.

[0130] Experiment 1: The recognition rates of six radar PRI modulation signals in six scenarios were compared using the method of the present invention. Fig.13 It can be seen that the proposed method can basically achieve a recognition accuracy of more than 90% for the six radar PRI modulation modes when the pulse loss ratio is less than 50% and the false pulse ratio is less than 40%. It can basically be correctly identified when the pulse loss ratio is less than 30% and the false pulse ratio is less than 24%, which verifies the effectiveness of the proposed method.

[0131] Experiment 2: Compare the algorithm of the present invention with the methods of CNN+LSTM, TCN and CNN algorithms, and perform performance analysis based on the simulation results. Fig.14 It can be seen that the recognition effect of the algorithm of the present invention is better than that of other classical neural network methods under high noise conditions. Compared with the CNN+LSTM method with better recognition effect, the training time of this method is 2944s, and the single sample test time is 9.4ms; the training time of the method of the present invention is 2523s, and the single sample test time is 8.5ms, which verifies the high efficiency of the designed new algorithm model.

Claims

1. A radar PRI modulation type recognition method based on neural network and multi-head attention mechanism, characterized in that: The following steps are involved: Step 1, simulate and generate six types of pulse-interval modulation radar signals, including fixed modulation signal, staggered modulation signal, jitter modulation signal, group-variable modulation signal, sinusoidal modulation signal and sliding modulation signal, perform noise processing on the above six types of pulse-interval modulation radar signals with different proportions of missing pulses and false pulses, and obtain the radar signal arrival time TOA sequence data set; Step 2: Perform a differential calculation on the radar signal TOA sequence to obtain the pulse repetition interval PRI sequence of the radar signal. Then, use a median filter to filter the noisy PRI sequence containing false pulses and missing pulses, and then divide all PRI sequences into a training set and a test set. Step 3, design a multi-scale convolution squeeze excitation neural network MSCSE module to fuse the extracted local features with different scales and different weights into a feature vector and input it into the next layer of neural network; Step 4: Build a radar PRI modulation recognition algorithm model based on neural network and multi-head attention mechanism. The model includes MSCSE module, bidirectional gated unit BiGRU network module and multi-head attention mechanism MHA module. The BiGRU network uses the context features of both the front and back directions in the sequence. The MHA module highlights the key information of the BiGRU network at different time steps, enhancing the algorithm model's ability to focus on important information in the data. Step 5: Send the generated PRI sequence training set into the algorithm model for model training, adjust the model's hyperparameters, update the model's weights and biases, and use the trained algorithm model to send the test sets under different noise conditions into the model for classification and recognition, so as to obtain the recognition rate of the radar signal PRI modulation mode under different noise conditions.

2. The radar PRI modulation type recognition method based on neural network and multi-head attention mechanism according to claim 1 is characterized in that: In step 1, six PRI modulated radar signals are simulated and generated. The time interval between the arrival of adjacent pulses is the pulse repetition interval PRI, which is defined by the function as follows: P={p1,p2,…,p N },p i =t i+1 -t i Where P is the PRI sequence, {t i ,i=1,2,…,N+1} is the arrival time of each pulse signal in the radar pulse train being analyzed, i is the data sequence number, N is the number of radar pulse repetition intervals contained in the radar pulse train, and there are many different working modes and modulation modes for PRI in different radar systems. In practical applications, due to the influence of noise, the PRI sequence has a high loss and spurious pulse rate. Random noise is added to the TOA sequence of the generated six PRI modulation signals, including pulse missing noise and pulse false noise. The missing pulse sequence can be expressed as: Among them, {p′ n } indicates the received TOA sequence T n The new PRI sequence when j pulses are missing from the i-th pulse to the i+j-1 pulse, the false pulse sequence can be expressed as: {p′ n ′} represents a new PRI sequence with j pulses added after the i-th pulse, T i is the interval between the last false pulse and the next true pulse of the ith pulse.

3. The radar PRI modulation type recognition method based on neural network and multi-head attention mechanism according to claim 2 is characterized in that: The PRI of different radar systems has a variety of different working modes and modulation modes, including the following 6 types of PRI modulation type signals: Fixed PRI modulation signals generally have a fixed value, and their fluctuation range does not exceed 1%; In the staggered PRI modulated signal, multiple PRI values ​​appear in a periodic sequence; In a jittered PRI modulated signal, the value of the PRI sequence jitters around a constant value. The jitter range is usually randomly selected between 3% and 30%, usually following a Gaussian distribution. The value of the slip modulation PRI sequence usually increases monotonically from an initial value until it reaches a maximum value, or decreases monotonically from a maximum value to a minimum value; The numerical variation law of the sinusoidal modulation PRI sequence has a shape similar to that of a sine function and appears periodically; Group-variable modulation is similar to staggered modulation, with several fixed values. The difference is that different PRI values ​​in group-variable modulation have a certain residence time, and the number of PRI residence arrays also changes periodically.

4. The radar PRI modulation type recognition method based on neural network and multi-head attention mechanism according to claim 1 is characterized in that: In step 2, the n },{p′ n The first-order difference calculation is performed on each inter-pulse modulation TOA sequence of ′} to obtain the PRI sequence, which is first normalized to avoid a large difference in the order of magnitude of different data and to speed up the calculation. The normalization process is expressed as: Among them, s max is the maximum value in the current sequence, s min is the minimum value in the current sequence. Then, each normalized PRI sequence is subjected to median filtering. The median filter can effectively suppress the noise caused by missing pulses and false pulses. For a window of size 2K+1, it is expressed as: x i =median{and i-K ,and i-K+1 ,…,and i+K } where x i It is the data point after filtering. The new PRI sequence data after median filter processing is divided into training set and test set. The ratio of training set to data set is 8:

2.

5. The radar PRI modulation type recognition method based on neural network and multi-head attention mechanism according to claim 4 is characterized in that: The total number of samples in the dataset is 60,000.

6. The radar PRI modulation type recognition method based on neural network and multi-head attention mechanism according to claim 1 is characterized in that: In step 3, a multi-scale convolution module MSCSE integrating the squeeze-excited attention mechanism is designed. First, four convolution kernels of different sizes, 1×1, 1×3, 1×5, and 1×7, are used to process the data. The model can capture features of different scales. Smaller convolution kernels focus on local fine-grained features, while larger convolution kernels can capture longer-range dependencies, avoiding information loss that may be caused by relying only on a single-scale convolution kernel. The calculation formula for a convolution kernel size of 1×k can be expressed as: Among them, a represents the ath channel under the convolution kernel operation, Represents the convolution weight of the ath channel of the convolution kernel, Indicates the corresponding offset, X t is a subsequence of the input data intercepted at time step t, is the corresponding output result. After the convolution process, the output result Y is obtained. k After batch normalization, the output of each convolutional layer is concatenated to obtain the total output Y of a multi-scale convolution. The concatenation operation formula is as follows: Y=Concat(Y1,Y3,Y5,Y7) Subsequently, through the squeezing and excitation process, the module can adaptively adjust the weights of the feature channels so that the network pays more attention to important features. First, in the squeezing process, the input feature map dimension is H*1*C, which represents the length, width and number of channels of the feature map respectively. Global average pooling is performed on each channel to compress the feature map of each channel into a value, namely 1*1*C, to obtain a global description of each channel. The operation expression is as follows: Among them, F sq Represents the squeeze operation, Y(i) represents the input feature of the i-th channel, followed by the excitation operation, which inputs the channel description vector into a small fully connected network, including two fully connected layers FC and an activation function Sigmoid, to generate the weight coefficient of each channel. The weight coefficient is the scale factor for adjusting the original feature map. The calculation process expression is as follows: ω c =F ex (Z c ,W)=Sigmoid(W2(W1Z c +b1)+b2) Among them, F ex represents the excitation operation, W1 and b1 are the weight and bias of the first fully connected layer, W2 and b2 are the weight and bias of the second fully connected layer, and finally, the learned weight coefficient is weighted to each channel feature, and the result is added to the input feature to get the final output. The calculation expression is as follows: Y'=ω c Y+Y Afterwards, the result is processed with the P-ReLU activation function. P-ReLU is an improved activation function that introduces learnable parameters based on the ReLU activation function to avoid the inability to learn effective features in the negative value area to enhance the expressiveness of the model. The calculation expression is as follows: X'=f(Y') Among them, α is a learnable parameter, and then the result X' processed by the P-ReLU activation function is sent to the next layer of the network.

7. The radar PRI modulation type recognition method based on neural network and multi-head attention mechanism according to claim 1 is characterized in that: In step 4, a radar PRI modulation recognition algorithm model is constructed. The preprocessed PRI sequence data first passes through two MSCSE layers to extract features at different time scales, and then passes through a pooling layer to downsample the features to reduce the data dimension while retaining important information to further improve the computational efficiency. Then, after passing through two BiGRU layers, the model can better learn the temporal relationship in the sequence and capture the context information. Subsequently, after passing through the multi-head attention layer model, the model can focus on the most important part of the PRI sequence, making information transmission more efficient. The data is flattened through the Flatten layer. Finally, the extracted features are integrated through a fully connected layer, and finally the classification results are output through the Softmax layer, thereby achieving accurate recognition of different types of PRI modulation.

8. The radar PRI modulation type recognition method based on neural network and multi-head attention mechanism according to claim 1 is characterized in that: In step 5, the generated PRI sequence training set is sent to the algorithm model for model training. The training process includes setting hyperparameters and updating network weights. Assuming that the final prediction result of the model is p q , the loss function used is the root mean square error function, and the calculation expression is as follows: Where B is the total number of batch samples, which is the same as the value of the hyperparameter Batch_size, and l q is the label value of the qth sample, p q is the predicted value of the qth sample, and RMSE is the loss function value, which is used for back-propagation to update the model weights.

9. The radar PRI modulation type recognition method based on neural network and multi-head attention mechanism according to claim 8 is characterized in that: The model parameters are set as follows: the optimizer uses Adam, the batch size is 128, the epochs is 50, and the Lr learning rate is 0.

001. After that, the trained prediction model is tested using the test set to verify that the model has a high prediction accuracy.

Images