A method for evaluating electronic interference effect by extracting radar parameter change amount

Abstract

The application discloses an electronic interference effect evaluation method for extracting radar parameter change quantity, and solves the problems that the existing interference effect evaluation technology considers single influencing factor, and the interference effect evaluation result is subjective and easy to misjudge. The steps of the application comprise the following steps: acquiring radar state characteristics by using intercepted radar data parameters; adopting a Bayesian neural network as an interference effect evaluation network; training the interference effect evaluation network; and quantitatively evaluating the interference effect. The application proposes an interference effect evaluation method which is more suitable for the existing complex electronic countermeasure environment, simplifies the complexity of the evaluation system, reduces the requirement for the training data quantity, and has higher interference effect evaluation accuracy.

Description

Technical Field

[0001] This invention belongs to the field of radar technology, and further relates to an electronic jamming effect evaluation method for extracting radar parameter changes in the field of radar jamming technology. This invention can be used in electronic warfare to accurately evaluate the jamming effect of the jammer. Background Technology

[0002] In electronic warfare, jamming effectiveness assessment is a crucial part of the jamming side's operational planning. It involves the qualitative or quantitative analysis of the impact on radar after jamming has been implemented. The assessment results serve as an important reference for timely adjustments to jamming strategies, thereby improving the effectiveness of radar jamming measures.

[0003] In their paper "An Online Evaluation Method for Jamming Effect Based on Radar State Changes" (Electronic Information Countermeasures Technology, 2016, 31(03): 42-46), Zhao Yaodong et al. proposed a method for online evaluation of jamming effects based on changes in radar operating state. The implementation scheme is as follows: First, the influencing factors of online evaluation of jamming effects based on the jammer's reconnaissance radar signals are analyzed. Then, by establishing a mapping relationship between changes in radar operating state and jamming effects, and combining radar attribute recognition and state recognition, the implementation process of the online evaluation method for jamming effects is given. Finally, the method is verified and illustrated by corresponding mathematical models and simulation examples. This method uses radar signals intercepted by the jammer to evaluate jamming effects, solving the problem that traditional offline evaluation methods struggle to obtain enemy radar performance parameters in real-world scenarios. However, this method still has two shortcomings. First, it judges the effectiveness of jamming measures based solely on changes in radar operating status. When the radar operating status remains unchanged but the anti-jamming method changes, the jamming effect assessment result will be misjudged. Second, the knowledge base of this method can only obtain a judgment on whether the jamming is effective, but cannot provide a more detailed description of the effectiveness of the jamming measures. This results in too little information available for jamming decision-making, making it difficult to further improve the effectiveness of jamming measures.

[0004] Xi'an University of Electronic Science and Technology proposed a radar jamming effect evaluation method in its patent application "A Radar Jamming Effect Evaluation Method, Device and Computer Equipment" (Application No.: 201910229296.7; Publication No.: CN 110082733 A). The method's implementation involves: first, analyzing the behavioral parameters of the target radar and the radar jamming scheme to establish a set of radar jamming factors such as jamming timing, jamming frequency, and jamming range; obtaining the jamming benefit value of each jamming factor through a pre-set method; then, obtaining the user-input score of the jamming factor; subjectively determining the weight vector of each factor using the analytic hierarchy process (AHP); finally, synthesizing the jamming benefit values ​​and weight vectors of each jamming factor to obtain a weighted jamming benefit matrix; and using the Top-Order Solution Approximation (TOPSIS) method to obtain the radar jamming scheme's jamming effect score. This method can achieve online jamming effect evaluation. However, its shortcomings include the fact that the evaluation results rely on user evaluations and expert experience to determine the weights of each evaluation indicator, making the evaluation results susceptible to subjective human factors and resulting in low reliability. Summary of the Invention

[0005] The purpose of this invention is to address the shortcomings of the prior art by proposing an electronic jamming effect evaluation method that extracts radar parameter changes. This method solves the problem that existing jamming effect evaluation techniques consider only one influencing factor and are prone to misjudgment of jamming effect evaluation results. It can also achieve an objective and accurate quantitative evaluation of jamming effects, ensuring the reliability of evaluation results in actual combat scenarios.

[0006] The specific approach to achieving this invention is as follows: This invention extracts parameters from radar data intercepted before and after interference, considering factors such as changes in radar behavior before and after interference, the use of anti-interference methods, and the suitability of the jamming measures taken by the jammer. It establishes a comprehensive set of interference effect evaluation parameters, reducing the probability of misjudgment and thus addressing the problem of existing interference effect evaluation technologies considering only a single influencing factor, leading to easy misjudgment of the evaluation results. This invention standardizes the interference effect evaluation parameters for different interference methods and varying state characteristics of enemy radars. Then, it generates a training set containing the interference effect evaluation parameters and results, constructs an interference effect evaluation network structure, and trains the network with relatively little training data to obtain an interference effect evaluation network model with good nonlinear mapping capabilities. This invention inputs the processed interference scheme evaluation parameter data into the trained evaluation network model to obtain the evaluation results, overcoming the problem of traditional methods being greatly affected by subjectivity.

[0007] The implementation steps of this invention are as follows:

[0008] Step 1: Obtain radar state characteristics using intercepted radar data parameters;

[0009] The jamming party extracts parameters including pulse amplitude, pulse carrier frequency, pulse width, and intra-pulse modulation characteristics from the intercepted radar signal data, identifies the state characteristics of the opponent's radar, and constructs radar state transition degree and identification probability index respectively.

[0010] Step 2: Generate a training set for the interference effect evaluation parameters;

[0011] A dataset of at least 50 jamming samples was selected. Each sample contained five jamming effect evaluation parameters: radar state transition degree and recognition probability, jammer frequency aiming accuracy, power-to-interference-to-signal ratio, and jamming pattern matching accuracy. Jamming experiments were performed on the jamming effect evaluation parameters of each sample to obtain the jamming effect evaluation value of that sample. The normalized jamming effect evaluation parameters of all samples and their corresponding jamming effect evaluation values ​​were combined to form a training set.

[0012] Step 3: Use a Bayesian neural network as the network for evaluating the interference effect;

[0013] Step 4: Train the interference effect evaluation network;

[0014] The training set is input into the interference effect evaluation network, and the weight values ​​of the network parameters are iteratively updated using the backpropagation algorithm until the objective function converges, thus obtaining the trained interference effect evaluation network.

[0015] Step 5: Quantitatively evaluate the interference effect;

[0016] The normalized evaluation parameters of the interference scheme to be evaluated are input into the trained interference effect evaluation network, and the interference effect evaluation results are output.

[0017] Compared with the prior art, the present invention has the following advantages:

[0018] First, this invention establishes a set of interference effect evaluation parameters, including state transition degree, identification probability, jamming frequency aiming degree, power interference-to-signal ratio, and jamming pattern matching degree. Furthermore, it standardizes the interference effect evaluation parameters for different radar behavior states and jamming patterns, overcoming the shortcomings of traditional methods in that the evaluation parameters are not comprehensive and the evaluation results are prone to misjudgment. This invention greatly improves the accuracy of electronic jamming effect evaluation.

[0019] Secondly, this invention utilizes a Bayesian neural network to fit the nonlinear relationship between the interference effect evaluation parameters and the interference effect, which significantly reduces the requirements for the amount of training data during network training. This effectively prevents overfitting of existing neural networks when there are too many indicator factors or too little training data. At the same time, it avoids the influence of human subjective factors on the evaluation results of existing evaluation algorithms, thus improving the reliability of interference effect evaluation. Attached Figure Description

[0020] Figure 1 This is a flowchart of the present invention;

[0021] Figure 2 This is a comparison chart of the simulation results of the present invention. Detailed Implementation

[0022] The present invention will now be further described with reference to the accompanying drawings and embodiments.

[0023] In this embodiment of the invention, it is assumed that the radar is powered on both before and after being interfered with. If the radar is powered off after being interfered with, the interference effect can be considered to be the best.

[0024] Reference Figure 1 The implementation steps of the embodiments of the present invention will be further described below.

[0025] Step 1: Obtain radar state characteristics using intercepted radar data parameters.

[0026] The jamming party extracts parameters including pulse amplitude, pulse carrier frequency, pulse width, and intra-pulse modulation characteristics from the intercepted radar signal data, identifies the state characteristics of the opponent's radar, and constructs radar state transition degree and identification probability index respectively.

[0027] The radar state characteristics refer to the identification results of radar behavior state and anti-jamming methods obtained by analyzing parameters extracted by the jammer from the intercepted radar signal data using a radar operating mode recognition algorithm.

[0028] The radar behavior states are: search, confirm, track, and lose track, totaling four states.

[0029] The anti-interference method is waveform pattern anti-interference, specifically including: frequency agility, carrier frequency agility, frequency modulation slope agility, orthogonal frequency division linear frequency modulation, phase coding, and frequency coding, totaling 6 measures.

[0030] In this embodiment of the invention, the radar's behavior before and after being interfered with, the anti-interference method, and the accuracy of radar state feature recognition are taken as known conditions.

[0031] Step 2: Generate a training set for the interference effect evaluation parameters.

[0032] This invention selects 50 interference samples to form a dataset. Each sample contains five interference effect evaluation parameters: radar state transition degree and recognition probability, jammer frequency aiming accuracy, power-to-interference-to-signal ratio, and interference pattern matching accuracy. Interference experiments are performed on the interference effect evaluation parameters of each sample to obtain the true interference effect evaluation value for that sample, with the value limited to between 0 and 1. All normalized sample interference effect evaluation parameters and their corresponding true interference effect evaluation values ​​are then combined to form a training set.

[0033] The state transition degree is a specific representation of the changes in the state characteristics of the radar before and after being interfered with. By analyzing the changes in radar behavior and anti-interference methods, and matching the interference effect knowledge base, the state transition degree is standardized into discrete points at equal intervals in the range of 1 to 10. The larger the discrete point value, the higher the interference effectiveness.

[0034] The recognition probability is obtained by multiplying the accuracy of the radar's state feature recognition before interference with the accuracy of the radar's state feature recognition after interference.

[0035] The frequency targeting accuracy of the jammer describes the degree of overlap between the jamming frequency and the radar's operating frequency. The higher the frequency targeting accuracy of the jammer, the better the jamming effect. The calculation formula is as follows:

[0036]

[0037] Among them, D f f represents the frequency targeting accuracy of the jammer. jl ~f jh f represents the frequency range in which the jammer operates. rl ~f rh Indicates the frequency range in which the radar operates.

[0038] The power-to-interference-to-signal ratio (HIRR) is obtained by dividing the power of the jamming signal by the power of the radar signal. The higher the HIRR, the lower the signal-to-interference ratio of the radar received signal, and the better the jamming effect.

[0039] The jamming pattern matching degree is used to characterize the jamming effect of different jamming patterns on radars under different behavioral states. The jamming pattern matching degree is standardized into discrete points at equal intervals in the range of 1 to 10. The larger the discrete point value, the more effective the jamming. The jamming effect matching degree of suppression jamming on search, confirmation, loss of tracking and tracking behavior states is 10, 9, 7 and 5, respectively. The jamming effect matching degree of deception jamming on search, confirmation, loss of tracking and tracking behavior states is 5, 7, 9 and 10, respectively.

[0040] The normalization of the interference effect evaluation parameters for the samples is performed by the following formula:

[0041]

[0042] Among them, b ij This represents the value of the j-th interference effect evaluation parameter corresponding to the i-th sample after normalization. j=1 represents the state transition degree, j=2 represents the recognition probability, j=3 represents the interference frequency targeting accuracy, j=4 represents the power-to-intrusion ratio, and j=5 represents the interference pattern matching degree. ij This represents the value of the interference effect evaluation parameter corresponding to the i-th sample, This represents the minimum value of the j-th interference effect evaluation parameter across all datasets. This represents the maximum value of the j-th interference effect evaluation parameter across all datasets.

[0043] Step 3: In this invention, a Bayesian neural network is used as the network for evaluating the interference effect.

[0044] The interference effect evaluation network consists of an input layer, a hidden layer, and an output layer connected in series. The number of nodes in the input layer is set to 5, the number of nodes in the hidden layer is 6, and the number of nodes in the output layer is 1. The transfer function of the interference effect evaluation network is the tansig kernel function.

[0045] Step 4: Train the interference effect evaluation network.

[0046] The training set is input into the interference effect evaluation network, and the weight values ​​of the network parameters are iteratively updated using the backpropagation algorithm until the objective function converges, thus obtaining the trained interference effect evaluation network.

[0047] The objective function is:

[0048]

[0049] Where F represents the objective function, k represents the index of the sampling point of the distribution function, k = 1, ..., n, n represents the number of sampling points of the distribution function, log(·) represents the logarithmic operation with the natural constant e as the base, q(·|·) represents the distribution function, and w (k) θ represents the weight of the data corresponding to the k-th sampling point of the distribution function, θ represents the parameter set of the control weight w that follows a Gaussian distribution, including the mean and standard deviation, p(·) represents the prior probability density function, p(·|·) represents the conditional probability density function, and X represents the total number of samples in the training set of the interference effect evaluation parameters.

[0050] The Gaussian prior distribution assumption of the weights w in the interference effect evaluation network is equivalent to introducing an L2 regularization term, which can prevent overfitting during network training when the training set size is too small or the network complexity is too high, thus reducing the requirement for the size of the training set.

[0051] Step 5: Quantitatively evaluate the interference effect;

[0052] Using the same normalization process as in step 2, the normalized evaluation parameters of the interference scheme to be evaluated are input into the trained interference effect evaluation network, and the interference effect evaluation results are output.

[0053] The effects of this invention will be further illustrated below with simulation experiments:

[0054] 1. Simulation experimental conditions:

[0055] The hardware platform for the simulation experiment of this invention is: Intel(R) Core i7-7700HQ CPU with a main frequency of 2.8GHz and 16GB of memory.

[0056] The software platform for the simulation experiment of this invention is: Windows 11 operating system and MATLAB R2021b.

[0057] The parameters of the training dataset, test dataset, and comparison sample dataset used in the simulation experiment of this invention are set as follows: the pulse repetition frequency range of the radar signal is 2-15KHz, the carrier frequency range is 2.5-3GHz, the bandwidth range is 2-20MHz, the pulse width range is 1-100us, and the power range is 4-18kW; the carrier frequency range of the jamming signal is 2-4GHz, the jamming bandwidth range is 5-30MHz, and the jamming signal power range is 2-10kW.

[0058] The training set for the interference effect evaluation network in the simulation experiment of this invention is set to 50 samples, the test set is set to 20 samples, and the sample data size for the comparison experiment is set to 12 groups. Each group of data contains 5 interference effect evaluation parameters and their corresponding real interference effect evaluation results, and the real interference effect evaluation results of each group of data have a large degree of discrimination.

[0059] The network configuration parameters for evaluating the interference effect in the simulation experiment of this invention are as follows: the maximum number of training iterations is set to 1000, the learning rate is set to 0.02, and the minimum training error is set to 0.000001.

[0060] 2. Simulation content and result analysis:

[0061] The interference effect evaluation simulation experiment of this invention uses the analytic hierarchy process (AHP) of this invention and existing technologies to evaluate the interference effect on 12 sets of input sample data, obtaining the evaluation results of 12 corresponding interference effects, and plotting the interference effect evaluation results as shown in the figure. Figure 2 As shown.

[0062] In simulation experiments, the existing technologies used refer to:

[0063] The existing analytic hierarchy process (AHP) refers to the interference effect evaluation method proposed by Tang Guangfu, An Hong, and others in "Evaluation of Collaborative Interference Effectiveness Based on AHP" (Electronic Information Countermeasures Technology, 2016, 31(4)).

[0064] The following is combined with Figure 2 The simulation results further illustrate the effects of the present invention.

[0065] Figure 2This is a comparison chart showing the interference effect evaluation results for each sample data obtained using the method of this invention and existing technologies. Figure 2 The horizontal axis in the graph represents the sample data sequence number. Figure 2 The vertical axis in the figure represents the magnitude of the interference effect evaluation result. Figure 2 The curves marked with star symbols represent the true value curves of the interference effect evaluation results for each sample data. Figure 2 The curves marked with circular symbols represent the curves of interference effect evaluation results calculated using the method of this invention for each sample data. Figure 2 The curves marked with the upper triangle symbol represent the results of interference effect evaluation calculated using the analytic hierarchy process (AHP) for each sample data.

[0066] Depend on Figure 2 As can be seen from the simulation results, the interference effect evaluation calculation results of the method of the present invention are close to the true values ​​under each group of sample data. However, the interference effect evaluation calculation results of the prior art analytic hierarchy process are affected by the subjective setting of evaluation parameter weights, resulting in a large deviation between the evaluation results and the true values.

[0067] To further demonstrate the effectiveness of the present invention, the coefficient of determination R is used. 2 As a parameter for evaluating the performance of interference effects, its value is between 0 and 1. The closer the value is to 1, the better the performance of the evaluation method. The calculation formula is as follows:

[0068]

[0069] Where ∑ represents the summation operation, t represents the index of the sample data, t = 1, ..., m, m represents the total number of sample data, m = 12, y t ' represents the predicted evaluation value of the interference effect corresponding to the t-th sample data, y t This represents the true value of the interference effect corresponding to the t-th sample data. This represents the mean of the true values ​​of the interference effect for all sample data.

[0070] The interference effect evaluation result of the method of the present invention can be calculated using the above formula. 2 =0.98, R² of the interference effect evaluation result of the analytic hierarchy process. 2 =0.73. Therefore, it can be seen that the method of this invention has superior performance in evaluating interference effects compared to the analytic hierarchy process (AHP), and is more suitable for evaluating interference effects in existing complex electronic warfare environments, yielding more objective and accurate evaluation results.

Claims

1. A method for evaluating the effectiveness of electronic jamming by extracting changes in radar parameters, characterized in that, A training set is generated to evaluate the interference effect, including parameters such as state transition degree, recognition probability, interference frequency targeting accuracy, power-to-intrusion ratio, and interference pattern matching accuracy. A Bayesian neural network is then used to fit the nonlinear relationship between the interference effect evaluation parameters and the interference effect. The steps of this evaluation method include the following: Step 1: Obtain radar state characteristics using intercepted radar data parameters: The jamming party extracts parameters including pulse amplitude, pulse carrier frequency, pulse width, and intra-pulse modulation characteristics from the intercepted radar signal data, identifies the state characteristics of the opponent's radar, and constructs radar state transition degree and identification probability index respectively. The state characteristics of the opposing radar refer to the analysis of parameters extracted from the intercepted radar signal data by the interfering party using a radar working mode recognition algorithm to obtain the identification results of radar behavior state and anti-interference method. The radar behavior state includes four states: search, confirmation, tracking, and loss of tracking. The anti-interference method is waveform pattern anti-interference. Step 2, generate a training set for interference effect evaluation parameters: A dataset of at least 50 jamming samples was selected. Each sample contained five jamming effect evaluation parameters: radar state transition degree and recognition probability, jammer frequency aiming accuracy, power interference-to-signal ratio, and jamming pattern matching accuracy. Jamming experiments were performed on the jamming effect evaluation parameters of each sample to obtain the true jamming effect evaluation value of the sample. The normalized jamming effect evaluation parameters of all samples and their corresponding true jamming effect evaluation values ​​were combined to form a training set. The radar state transition degree is a specific representation of the changes in radar state characteristics before and after interference. By analyzing the changes in radar behavior state and anti-interference methods, and matching the interference effect knowledge base, the state transition degree is standardized into discrete points at equal intervals in the range of 1 to 10. The larger the discrete point value, the higher the interference effectiveness. The recognition probability is obtained by multiplying the accuracy of the radar's state feature recognition before interference with the accuracy of the radar's state feature recognition after interference. The jamming pattern matching degree is used to characterize the jamming effect of different jamming patterns on radar under different behavioral states. The jamming pattern matching degree is standardized into discrete points at equal intervals in the range of 1 to 10. The larger the discrete point value, the more effective the jamming. The jamming pattern matching degree of suppression jamming on search, confirmation, loss of tracking and tracking behavior states is 10, 9, 7 and 5 respectively, and the jamming pattern matching degree of deception jamming on search, confirmation, loss of tracking and tracking behavior states is 5, 7, 9 and 10 respectively. Step 3: Use a Bayesian neural network as the network for evaluating the interference effect; Step 4, train the interference effect evaluation network: The training set is input into the interference effect evaluation network, and the weight values ​​of the network parameters are iteratively updated using the backpropagation algorithm until the objective function converges, thus obtaining the trained interference effect evaluation network. Step 5: Quantitatively evaluate the interference effect; Using the same normalization process as in step 2, the normalized evaluation parameters of the interference scheme to be evaluated are input into the trained interference effect evaluation network, and the interference effect evaluation results are output.

2. The electronic jamming effect evaluation method for extracting radar parameter changes according to claim 1, characterized in that, The normalization operation described in step 2 is performed by the following formula: ； in, Represents the normalized i-th The corresponding sample The values ​​of the interference effect evaluation parameters Time indicates the degree of state transition. Time represents the probability of recognition. The time indicates the accuracy of the jamming frequency aiming. The time indicates the power-to-interference ratio. Time indicates the degree of interference style matching. Indicates the first The corresponding sample One interference effect evaluation parameter value, Indicates the first The minimum value of each interference effect evaluation parameter across all datasets. Indicates the first The maximum value of each interference effect evaluation parameter across all datasets.

3. The electronic jamming effect evaluation method for extracting radar parameter changes according to claim 1, characterized in that, The interference effect evaluation network described in step 3 consists of an input layer, a hidden layer, and an output layer connected in series. The number of nodes in the input layer is set to 5, the number of nodes in the hidden layer is 6, and the number of nodes in the output layer is 1. The transfer function of the interference effect evaluation network is the tansig kernel function.

4. The electronic jamming effect evaluation method for extracting radar parameter changes according to claim 1, characterized in that, The objective function mentioned in step 4 is: ； in, Describe the objective function. This indicates the index of the sampling point of the distribution function. , This indicates the number of sampling points for the distribution function. Represented by natural constant Logarithmic operations with base 0. Represents the distribution function. Represents the distribution function of the th The weight of the data corresponding to each sampling point Indicates control weight A parameter set that follows a Gaussian distribution, including the mean and standard deviation. This represents the prior probability density function. This represents the conditional probability density function. This represents the total number of samples in the training set for the interference effect evaluation parameters.

Images