Active interference pattern recommendation method for unknown radar signal

Abstract

The invention provides an active interference pattern recommendation method for an unknown radar signal. The method comprises the following steps: acquiring parameter information of a known target radar signal; selecting parameters effective to target signal interference; constructing a test database and an interference experience library of known radar signals according to the selected parameters; according to test database information induction and summarization, extracting parameter information, and establishing an interference pattern experience library data sample table; vectorizing data samples in the experience library, and constructing a sample matrix; obtaining parameter information of an unknown radar target signal, extracting characteristic parameters corresponding to the target signal based on experience library parameters, and performing vectorization representation; performing similarity calculation on the target signal and the data in the experience library to obtain similarity weights with all the data; and selecting the interference pattern of the empirical data with the maximum weight as the optimal interference pattern of the unknown target for recommendation. According to the invention, more convenient and accurate interference styles are provided for decision makers.

Description

Technical Field

[0001] The present application relates to the field of electronic countermeasure technology, and in particular to a method for recommending active jamming patterns of unknown radar signals. Background Art

[0002] In modern warfare, various weapons and equipment are increasingly using and relying on radio electronic technology, especially radar systems in weapons and equipment. Radar is the most important equipment in the information battlefield and weapon system for target information acquisition, command and control, precision guidance and firepower strike combat system. Therefore, in today's information warfare, how to destroy the enemy's radar system and make it lose the ability to obtain information or obtain untrue information is the most important issue and the key to victory in the war.

[0003] With the continuous development of radar technology, the radar working mode has evolved from a fixed one to a multi-working mode that can adaptively adjust the working parameters according to the changes in the surrounding environment and task requirements. Therefore, for traditional electronic countermeasure equipment, when the enemy threat target is acquired by the operator, it is necessary to select an effective interference style based on the operator's personal theoretical knowledge and combat experience, through intuitive analysis and judgment, and make corresponding interference decisions on the threat target signal to achieve the purpose of destroying the enemy's reconnaissance. This traditional interference decision-making method based on manual operation cannot quickly and effectively interfere with multiple targets and tasks in a short period of time in actual warfare. It takes too long and the interference effect cannot be guaranteed. Therefore, this method is difficult to meet the current war needs in complex and changing environments.

[0004] For most of the current electronic countermeasure equipment, when the received target signal fails to match the equipment radar recognition library, it is considered an unknown signal. The operator is required to select a possible effective jamming pattern for it based on his theoretical knowledge and experience. This method requires high professional skills of the operator and is time-consuming. Therefore, in order to solve this problem, a method is proposed that can automatically recommend effective jamming patterns for unknown target radar signals. Summary of the invention

[0005] The present application provides a method for recommending active jamming patterns for unknown radar signals, which can be used to solve the technical problems of the prior art that are time-consuming and require high professional skills.

[0006] The present application provides a method for recommending an active jamming pattern for an unknown radar signal, the method comprising:

[0007] Step 1: Obtain the parameter information, test environment data, test process and result data of the known target radar signal, record and store all parameter information, including target batch number, true azimuth, heading angle, pitch, radio frequency, radio frequency type, repetition period, repetition type, pulse width, pulse width type, target distance, RCS, amplitude, radiation source type, attribute, scanning type, scanning period, bandwidth, pulse information, reconnaissance interference time, ship radar power-on status, interference style, transmission power, working mode, interference onset time, interference effect, polarization mode, effective interference time, electromagnetic environment background, electromagnetic compatibility control status, radar model;

[0008] Step 2: According to the known radar signal parameter information, select the parameters that are effective for interfering with the target signal based on the principle of electronic countermeasures; the effective parameters include radio frequency range F, radio frequency type T, agile range R, agile points N, repetition type Y, repetition range S, stagger / jitter points M, pulse width range W, target distance D, intra-pulse modulation parameter P, technical system C, and anti-interference measures G;

[0009] Step 3: Construct a test database of known radar signals and an interference experience database based on the selected parameters; determine the parameters and structural rules of each database;

[0010] Step 4: Summarize and summarize the test database information, extract parameter information, and establish a sample table of interference pattern experience database data;

[0011] Step 5: Vectorize the data samples in the experience database and construct a sample matrix;

[0012] Step 6: Obtain parameter information of unknown radar target signals, extract characteristic parameters corresponding to target signals based on empirical database parameters, and perform vector representation;

[0013] Step 7: Calculate the similarity between the target signal and the data in the experience database to obtain the similarity weight with all the data;

[0014] Step 8: According to the similarity weight, select the interference pattern of the empirical data with the largest weight as the optimal interference pattern for the unknown target for recommendation.

[0015] Under the premise of ensuring the correctness of known test data, the present invention can solve the errors and time loss existing in the manual selection of interference patterns in actual combat, provide decision makers with more convenient and accurate interference patterns, and realize automatic interference pattern recommendation for unknown radar target signals. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 The present invention is a flowchart of a method for recommending an active jamming pattern for an unknown radar signal.

[0017] Figure 2This is a structural diagram of the basic rule library of the present invention. DETAILED DESCRIPTION

[0018] In order to make the objectives, technical solutions and advantages of the present application clearer, the implementation methods of the present application will be further described in detail below with reference to the accompanying drawings.

[0019] The following first introduces the embodiments of the present application in conjunction with the accompanying drawings.

[0020] The present invention provides a method for recommending an active jamming pattern for an unknown radar signal, comprising the following steps:

[0021] In step 1, the target radar information is obtained, and relevant characteristic parameters are extracted and recorded, including: target batch number, true azimuth, heading angle, pitch, radio frequency, radio frequency type, repetition period, repetition type, pulse width, pulse width type, target distance, RCS, amplitude, radiation source type, attribute, scan type, scan period, bandwidth, intra-pulse information, etc.; the test environment data, test process and result data are obtained, and parameter information including the reconnaissance interference time of the target, the power-on status of the ship's radar, the interference style, the transmission power, the working mode, the interference onset time, the interference effect, the polarization mode, the effective interference time, the electromagnetic environment background, the electromagnetic compatibility control status, the radar model, etc. is extracted; all the above parameter values ​​that can be obtained are extracted according to the target signal, and for the parameter values ​​that cannot be obtained, the default blank value is filled with NaN; some parameters including the test environment, process and results are automatically recorded according to the test process or entered after the test is completed.

[0022] In step 2, based on the principle of electronic countermeasures, effective features for interfering with the target signal are extracted, including parameters such as RF range F, RF type T, agile range R, agile point number N, repetition type Y, repetition range S, staggered / jitter point number M, pulse width range W, target distance D, intra-pulse modulation parameter P, technical system C, and anti-interference measures G; the acquired feature parameters are used to construct a target type collection E, which is expressed as E={F, T, R, N, Y, S, M, W, D, P, C, G}.

[0023] In step 3, a test database, an interference experience database and its structural rules are constructed;

[0024] According to the parameters set in step 1, each test data record is extracted and recorded in the test database to build a test database. Each test data has one or more effective interference patterns matching it, denoted as K i ={A1, A2, A3, ..., A m}, where K i represents the i-th test data, A m is the mth characteristic parameter in step 1, there are m characteristic parameters in total, K i ∈[Z1,Z2,…,Zt ], where Z t represents the effective interference pattern t corresponding to the i-th test data; K = [K1, K2, ..., K l ], K represents the entire test database, there is a total of 1 test data, and then the data in the test database are summarized; let E i = {X1, X2, ..., X p}, where E i represents the i-th piece of experience data in the interference experience library, Xp represents the p-th characteristic parameter in the i-th piece of experience data in the interference experience library, and E i ∈[Z1,Z2,…,Z t ],Z t represents the effective interference pattern t corresponding to the i-th piece of empirical data;

[0025] Let X i =[x1:x n ],[x1:x n ] represents the interval range or data set of the data contained in the i-th characteristic parameter. According to the frequency change law, repetition change law, pulse width size or technical system of the test data, the interference pattern is combined and summarized. The data with the same interference pattern are unified and combined for analysis. According to the n features in the experience library, the data with similar or the same type of features are summarized to form experience data with different features. E=[E1,E2,...,E q ]Build an interference experience library to provide a reference for unknown signals. The following conditions must be met during the summary from the test database to the interference experience library:

[0026]

[0027] The specific test database parameters are shown in Table 1, and the empirical database parameters are shown in Table 2.

[0028] Table 1 Test data storage format

[0029]

[0030]

[0031] Table 2 Experience database storage format table

[0032]

[0033] In step 4, the interference experience database data sample table is established. The data is extracted and analyzed according to the inductive summary method in step 3 and stored in the database. Each piece of experience data has one or more corresponding interference patterns. The signals with the same RF type, repetition type and pulse width type in the test database are analyzed, and the interference patterns corresponding to each piece of data are merged and summarized. For example, there are m pieces of test data, one of which is K. i RF Type A i6 , Heavy Week Type A i8 , pulse width A i10 All are fixed, RF point set A i5 , double week point set A i7 , pulse width point set A i9 , the interference pattern is Z1, and the other data K i+j RF Type A (i+j)6 , Heavy Week Type A (i+j)8 , pulse width A (i+j)10 They are also fixed, and the radio frequency point set A (i+j)5 , the set of multiple points is A (i+j)7 , the pulse width point set is A (i+j)9 , the interference pattern is also Z1, and the radio frequency ranges of the two data are in the same band, then the two data are summarized as one empirical data and recorded in the interference experience library, where Fi = [A i5， A (i+j)5 ], T i For fixed, S i =[A i7 , A (i+j)7 ],Y i is fixed, W i =[A i10 , A (i+j)10 ], the interference pattern is Z1;

[0034] When the RF type, repetition type and pulse width type are consistent and the parameters are not much different, but the interference patterns are inconsistent, different patterns are included in the three interference patterns in the empirical data. When the RF type is agile, the range of the empirical data is adjusted according to the maximum and minimum values ​​of two or more data. Generally, the range of 300MHz, 500MHz and 1GHz can be set for distinction. Similarly, other types of signals are combined and summarized in the same way.

[0035] In step 5, the experience data samples in the interference experience library are vectorized to construct a sample matrix.

[0036] Vectorization is performed by encoding as shown in Table 3 or by using machine learning methods for vectorization encoding, such as one-hot encoding, or sparse encoding is implemented by using CountVectorizer in the machine learning scikit-learn library, as shown in Table 4. The constructed sample matrix is:

[0037]

[0038] Table 3 Encoding specifications of feature vectors

[0039]

[0040]

[0041] Table 4 Encoding specifications of feature vectors (machine learning method)

[0042]

[0043] In step 6, for the received unknown radar target signal, the parameters corresponding to the target are extracted according to the features in the interference experience library, and the parameters are vectorized by encoding according to the corresponding encoding specification in step 5 to obtain the vectorized representation of the unknown target.

[0044] In step 7, the similarity between the target radar signal and the experience data in the interference experience library is calculated to obtain the similarity weight between the target signal and all its experience data. The specific method is:

[0045] Assume that the unknown radar signal is x1=(x 11 , x 12 , …, x 1p ) and a certain empirical data in the known interference pattern database is x2=(x 21 , x 22 , …, x 2p ), where the means of x1 and x2 are and The similarity weight calculation method between the two is:

[0046]

[0047] d represents the similarity between two data vectors, and its value range is [-1,1]. The larger the value, the more similar they are. When d is equal to 0, the two vectors are unrelated; when d∈[0.8,1.0], they are extremely strongly correlated, d∈[0.6,0.8], they are strongly correlated, d∈[0.4,0.6], they are moderately correlated, d∈[0.2,0.4], they are weakly correlated, and d∈[0,0.2], they are weakly correlated or not correlated.

[0048] In step 8, the signal with the largest similarity weight is selected as the optimal interference pattern for recommendation. The specific method is as follows:

[0049] According to the similarity weights obtained by calculating the similarity between the target radar signal and all the known experience data in the interference experience library in step 7, they are sorted in descending order to obtain the similarity weight vector w = [d1, d2, d3, ..., d q ].

[0050] Step 8: Based on the similarity weights in step 7, recommend effective interference patterns in descending order of weights. The weights correspond to the empirical data one by one, and each piece of empirical data has a corresponding interference pattern. Therefore, the corresponding interference pattern can be directly recommended according to the weights as the most effective interference pattern for unknown radar signals.

[0051] The above-described embodiments of the present application do not constitute a limitation on the protection scope of the present application.

Claims

1. A method for recommending active jamming patterns for unknown radar signals, characterized in that: The method comprises: Step 1: Obtain parameter information of known target radar signals, test environment data, test process and result data, and record and store all parameter information; Step 2: Based on the known radar signal parameter information and the principle of electronic countermeasures, select the parameters that are effective in interfering with the target signal; Step 3: Construct a test database of known radar signals and an interference experience database based on the selected parameters; determine the parameters and structural rules of each database; Step 4: Summarize and summarize the test database information, extract parameter information, and establish a sample table of interference pattern experience database data; Step 5: Vectorize the data samples in the experience database and construct a sample matrix; Step 6: Obtain parameter information of unknown radar target signals, extract characteristic parameters corresponding to target signals based on empirical database parameters, and perform vector representation; Step 7: Calculate the similarity between the target signal and the data in the experience database to obtain the similarity weight with all the data; Step 8: According to the similarity weight, select the interference pattern of the empirical data with the largest weight as the optimal interference pattern for the unknown target for recommendation.

2. The method according to claim 1, characterized in that: The parameter information includes target batch number, true azimuth, heading angle, pitch, radio frequency, radio frequency type, repetition period, repetition type, pulse width, pulse width type, target distance, RCS, amplitude, radiation source type, attribute, scan type, scan period, bandwidth, intra-pulse information, reconnaissance interference time, ship radar power-on status, interference style, transmission power, working mode, interference onset time, interference effect, polarization mode, effective interference time, electromagnetic environment background, electromagnetic compatibility control status, and radar model.

3. The method for recommending active jamming patterns of unknown radar signals according to claim 1, characterized in that: Step 2: The effective parameters include RF range F, RF type T, agile range R, agile points N, repetition type Y, repetition range S, staggered / jitter points M, pulse width range W, target distance D, intra-pulse modulation parameter P, technical system C, and anti-interference measures G; obtain the interference-effective parameters and construct the target type collection E, expressed as E = {F, T, R, N, Y, S, M, W, D, P, C, G}.

4. The method for recommending active jamming patterns of unknown radar signals according to claim 1, characterized in that: Construct a test database of known radar signals and an experience database of interference based on the selected parameters; determine the parameters and structural rules of each database, including: Construct a test database, each test data has one or more matching effective interference patterns, denoted as K i ={A1, A2, A3, ..., A m }, where K i represents the i-th test data, A m is the mth characteristic parameter in step 1, there are m characteristic parameters in total, K i ∈[Z1,Z2,…,Z t ], where Z t represents the effective interference pattern t corresponding to the i-th test data; K = [K1, K2, ..., K l ], K represents the entire test database, there is a total of 1 test data, and then the data in the test database are summarized; let E i = {X1, X2, ..., X p }, where E i represents the i-th piece of experience data in the interference experience library, Xp represents the p-th characteristic parameter in the i-th piece of experience data in the interference experience library, and E i ∈[Z1,Z2,…,Z t ],Z t represents the effective interference pattern t corresponding to the i-th piece of empirical data; Let X i =[x1:x n ],[x1:x n ] represents the interval range or data set of the data contained in the i-th characteristic parameter. According to the frequency change law, repetition change law, pulse width size or technical system of the test data combined with the interference style, the data with the same interference style are unified and summarized. According to the n features in the experience library, the data with similar or the same type of features are summarized to form experience data with different features. E = [E1, E2, ..., E q ]Build an interference experience library to provide a reference for unknown signals. The following conditions must be met in the process of summarizing from the test database to the interference experience library:

5. The method for recommending active jamming patterns of unknown radar signals according to claim 1, characterized in that: In step 5, the experience data samples in the interference experience library are vectorized to construct a sample matrix, including: Vectorization is performed by encoding as shown in Table 3 or vectorization encoding is performed by using a machine learning method as shown in Table 4; the constructed sample matrix is: Table 3 Encoding specifications of feature vectors Table 4 Encoding specifications of feature vectors Table 4 is constructed through machine learning.

6. The method for recommending active jamming patterns of unknown radar signals according to claim 1, characterized in that: Step 7: Calculate the similarity between the target signal and the data in the experience database to obtain the similarity weights with all the data, including: Calculate the target radar signal x1 = (x 11 , x 12 , …, x 1p ) and the experience data x2 in the interference experience library = (x 21 , x 22 , …, x 2p ) is determined by the following method: d represents the similarity between two data vectors, and its value range is [-1,1]. The larger the value, the more similar they are. When d is equal to 0, the two vectors are unrelated. When d∈[0.8,1.0], they are extremely strongly correlated, d∈[0.6,0.8], they are strongly correlated, d∈[0.4,0.6], they are moderately correlated, d∈[0.2,0.4], they are weakly correlated, and d∈[0,0.2], they are weakly correlated or have no correlation.

7. The method for recommending active jamming patterns of unknown radar signals according to claim 1, characterized in that: According to the similarity weight, the interference pattern of the empirical data with the largest weight is selected as the optimal interference pattern for the unknown target for recommendation, including: sorting according to the similarity weight to obtain the similarity weight vector w = [d1, d2, d3, ..., d q ], recommend effective interference styles according to the order of weights from large to small.

Images