Low-interception signal concealment performance evaluation method and system, medium and program product

Abstract

The invention discloses a low-interception signal concealment performance evaluation method and system, a medium and a program product, and belongs to the field of anti-interference communication. Inputting various parameters of the signal, and calculating specific data according to nine indexes of different criterion layers; an analytic hierarchy process model is established, the concealment performance is researched as a target layer, different domains are divided into a criterion layer, an index layer is nine different indexes of each domain, and subjective weight coefficients are calculated; creating an entropy weight method model, improving a standardization formula, and introducing a correlation coefficient to calculate an objective weight of an index; the subjective and objective weights are balanced based on the combined weighting thought of the game theory, and a comprehensive index weight coefficient is obtained; and a good and inferior solution distance is adopted as a sorting method, a weighting matrix is obtained according to the comprehensive index weight coefficient and the normalized signal data, the relative fitting degree of each scheme is calculated, and sorting is performed according to the magnitude of the fitting degree to obtain an evaluation result. According to the method, measurement and comparison of the low-interception signal concealment performance are realized.

Description

Technical Field

[0001] The present invention belongs to the technical field of anti-interference communication, and in particular relates to a method, system, medium and program product for evaluating the concealment performance of a low intercept signal. Background Art

[0002] Covert communication technology is the core supporting technology in the field of communication electronic countermeasures. It is necessary to ensure the reliability of information transmission and the anti-interception and identification capabilities of signals at the same time. Generally, the detection of targets is achieved by intercepting or detecting weak signals. Low intercept probability signals adjust the signal waveform structure to reduce the detection energy efficiency ratio of the signal, while ensuring that even if the signal is detected, it cannot be effectively intercepted and interfered. Therefore, low interception signals have good electronic countermeasure capabilities. Commonly used signal systems include spread spectrum signals, burst signals, frequency hopping and time hopping signals, and large time-bandwidth product signals.

[0003] The concealment performance of low-intercept signals is the focus of attention of both parties in electronic countermeasure scenarios. Conducting a concealment performance evaluation of a signal can measure the concealment effectiveness before the signal is transmitted, select a suitable waveform framework, and predict the anti-interception results; at the same time, it can provide a basis for the interception party to optimize the interception receiver, so that the interception party can obtain greater information at a lower cost. Traditional concealment performance evaluation indicators mainly revolve around interception factors, and it is necessary to build a scenario model of transmitting first and then intercepting. Basic conditions need to be preset for specific field models. The analysis is not comprehensive enough and has limited adaptability. There is also a lack of exploration of parameters from the multi-dimensional perspective of the signal itself.

[0004] The existing literature proposes a stealth performance evaluation method for radar signal waveform domain (Yang Chengxiu, Wang Qianzhe, Peng Weidong, et al. Radar waveform domain LPI performance evaluation method based on IFA-HFS [J]. Journal of Beijing University of Aeronautics and Astronautics), which applies the improved firefly algorithm to the hesitant fuzzy set evaluation method, extracts five intra-pulse and inter-pulse indicators, and simulates and compares four different radar signals.

[0005] A radar signal low acquisition performance evaluation method combining group generalized intuitionistic fuzzy sets with subjective and objective weights (Wu Hua, et al. "Radar LPI performance evaluation method based on improved G-GIFSS algorithm." Systems Engineering and Electronic Technology 39.6 (2017): 1256-1260.), selects six indicators from three levels to calculate the membership.

[0006] Evaluation method of communication countermeasure combat effectiveness based on cloud model(L.Zhu,D.Wang,L.Jiang,Z.Qi,X.Chen,andY.Wang,in 2022IEEE International Conference on Signal Processing,Communications and Computing(ICSPCC),2022,pp),Aiming at the shortcomings of the traditional hierarchical analysis method that the weight vector does not reflect the randomness and fuzziness of expert judgment, a hierarchical analysis method improved by cloud model is proposed.

[0007] A ship equipment reliability evaluation model based on bidirectional normalized projection and improved TOPSIS technology (Jia Xiaoping, Jia Baozhu) proposed a bidirectional normalized projection distance to address the inadequacies and defects of the traditional calculation method of the distance between good and bad solutions. The model has a larger range of priority coefficient values ​​and solves the evaluation ranking problem under large samples.

[0008] The above-mentioned low intercept signal concealment performance evaluation method has the following problems: the number of selected indicators is too small and the theoretical basis is lacking, the relationship analysis between indicators is unclear, and the measurement of the signal itself is not comprehensive enough. The improvement of the evaluation method is not in-depth enough, and the comparison of the advantages and disadvantages before and after the improvement is not highlighted. Summary of the invention

[0009] The object of the present invention is to provide a method, system, medium and program product for evaluating the concealment performance of a low-intercept signal, so as to evaluate the concealment performance of a low-intercept signal.

[0010] The purpose of the present invention is achieved through the following technical solutions:

[0011] A method for evaluating the concealment performance of a low intercept signal comprises the following steps:

[0012] Step 1: Select low interception signals, calculate various indicators in different domains for the selected signals according to the indicator layer diagram, construct a data matrix and calculate according to the improved standardized formula;

[0013] Step 2: Create an improved AHP model, study the concealment performance as the target layer, divide different domains into the criterion layer, and the indicator layer contains a total of nine indicators in different domains, and then calculate the subjective weight coefficient;

[0014] Step 3: Create an entropy weight method model, improve the standardized formula and introduce the correlation coefficient to calculate the objective weight of the indicator;

[0015] Step 4: Balance the subjective and objective weights based on the combined weighting idea of ​​game theory to obtain a comprehensive indicator weight coefficient;

[0016] Step 5: Use the distance between good and bad solutions as the sorting method, obtain the weighted matrix based on the comprehensive indicator weight coefficient and the normalized signal data, calculate the relative fit of each solution, and sort by fit to get the evaluation result.

[0017] Furthermore, the index hierarchical diagram in step 1 includes an index layer, a criterion layer, a decision layer and a target layer, and the type and modulation parameters of the selected signal are input; the index vector of each signal is obtained as Γ=[a 1 ,a 2 ,b,c 1 ,c 2 ,d 1 ,d 2 ,d 3 ,d 4 ], construct the data matrix D = [Γ 1 ,Γ 2 ,Γ 3 ,…,Γ m ] T .

[0018] Furthermore, the indicator layer specifically includes:

[0019] Power spectrum exceeds value a 1 Defined as: signal power spectrum P I (f) Exceeding the mean of the noise power spectrum The ratio of the part to the power spectrum length is defined as:

[0020]

[0021] in, is the mean function, N l This indicator measures the ability of the signal power spectrum to hide in noise;

[0022] Power spectrum normalized deviation a 2 It is defined as: taking the sum of multiple peaks in the central frequency band of the signal power spectrum density as the benchmark, and comparing the difference with the white noise power spectrum benchmark, the normalized deviation is obtained:

[0023]

[0024] Among them, ζ can take a positive value, generally 20, P i (f) is the normalized power spectrum of white noise, P fmax (f) is the normalized power spectrum of the signal; the index deviation a 2 Measures the number and distribution of signal power spectrum peaks;

[0025] The instantaneous pulse feature b is to normalize the signal amplitude; extract its instantaneous amplitude, instantaneous frequency and instantaneous phase, and calculate the variance of each sequence as σ a , σ f and Assume a monotonically decreasing function cos(arctan(σ f )) represents the concealment effectiveness. The index range is (0-1). The larger the value, the better the stealth effect. The index is determined as:

[0026]

[0027] Where N is the signal length, σ a , σ f and represent the variance of instantaneous amplitude, instantaneous frequency and instantaneous phase respectively;

[0028] The fractal dimension of the indicator CWD is c 1 =D B +D I , where D B With D I represent the box dimension and information dimension respectively;

[0029] The box dimension is defined as: D B =-lnN l / lnδ, where δ is the grid side length of each box, N l is the number of boxes needed to cover the frequency spectrum object;

[0030] Information Dimension D I Defined as: Where P i is the probability of an element falling into the box, P i lgP i It is the calculation of information entropy;

[0031] Index discrete instantaneous time bandwidth product c 2 The formula is The discrete time-bandwidth product is defined as:

[0032]

[0033] in, is the time-frequency function of the denoised signal, To find the mean function, To find the variance function; They are unit bandwidth and unit time respectively;

[0034] The fractal dimension index of the bispectral slice is: d 1 =DI ×2-(D B -1); D I With D B They are the box dimension and information dimension of the bispectral slice, respectively;

[0035] The bispectral diagonal integral index is: d 2 =D I ×2+10×(D B -1); D I With D B They are the box dimension and information dimension of the diagonal integral of the bispectrum, respectively;

[0036] Cycle spectrum index where Y W (p,q) and V Noise (p,q) are the cyclic spectra of Gaussian white noise respectively;

[0037] Cepstrum index d 4 =D B +10×D I , the cepstrum of the signal is The box dimension and information dimension are used to characterize the periodicity of the signal cepstrum.

[0038] Furthermore, the step 2 constructs a hierarchical analysis method model, and the indicators of each layer are scored by multiple experts. The weight vector obtained by improving the optimal transfer matrix and the cloud model includes the criterion layer w Crit =[ω A ,ω B ,ω C ,ω D ], solution layer The subjective weight of each solution level indicator relative to the highest target level in the improved analytic hierarchy process is calculated as follows:

[0039]

[0040] Furthermore, in step 3, an entropy weight method model is constructed, and the correlation coefficient is introduced to calculate the index conflict degree. The standardization formula is improved on the basis of range standardization, and the minimum value is scaled by the inverse tangent function, where γ takes a larger positive value:

[0041]

[0042] Calculate the correlation coefficient matrix between indicators; the Kendall correlation coefficient calculation formula between two indicators x and y is:

[0043]

[0044] Among them, ρ ijIt represents the correlation coefficient between the i-th index and the j-th index. C represents the number of pairs of elements with consistency for the indicators x and y, and D represents the number of pairs of elements with inconsistency. The objective weight W of the entropy weight method is calculated. φ 。

[0045] Further, step 4 constructs a comprehensive weight based on game theory. The following linear equations are obtained:

[0046]

[0047] The comprehensive weight is obtained by solving. where are the linear coefficients after normalization respectively.

[0048] Further, step 5 uses the distance between the optimal and worst solutions as the sorting method. First, the weighted sample matrix is obtained as X = Y·W. * , from which the positive and negative ideal solutions are obtained as follows:

[0049]

[0050] The distances from each scheme to the positive and negative ideal solutions are calculated. Finally, the relative fitness is obtained as: The schemes are sorted according to the size of the fitness to obtain the final evaluation result.

[0051] Further, a method for evaluating the concealment performance of low-intercept signals further includes the index selection principles in four criterion domains: the frequency domain, the waveform domain, the time-frequency domain, and the spectral analysis. Among them, the fractal dimension-based CWD time-frequency transformation dimensionality reduction processing, and the index calculation methods such as the fractal dimension of the bispectrum slice and the bispectrum diagonal integral are created.

[0052] A computer device / equipment / system includes a memory, a processor, and a computer program stored on the memory. The processor executes the computer program to implement the steps of a method for evaluating the concealment performance of low-intercept signals.

[0053] A computer-readable storage medium stores a computer program / instructions. When the computer program / instructions are executed by a processor, the steps of a method for evaluating the concealment performance of low-intercept signals are implemented.

[0054] A computer program product includes a computer program / instructions. When the computer program / instructions are executed by a processor, the steps of a method for evaluating the concealment performance of low-intercept signals are implemented.

[0055] The beneficial effects of the present invention are as follows:

[0056] The present invention proposes a multi-domain comprehensive low-interception signal concealment performance evaluation method, which realizes the effective measurement of the concealment effectiveness of various types of signals, and solves the problems of unclear indicators and unclear results caused by the single traditional concealment performance evaluation method and the evaluation criteria relying only on cyclic spectrum, power spectrum and inverse spectrum. The method proposed in the present invention divides the indicator modules according to the three stages of signal processing, comprehensively selects multiple indicator domains, and improves the subjective and objective weighting method. The cloud model is introduced to solve the multi-expert collaboration problem of the hierarchical analysis method, and the problem of repeated consistency testing in the hierarchical analysis method is solved by the optimal transfer matrix. The entropy weight method is improved by using the correlation coefficient to calculate the conflict degree between different indicators. The superior and inferior solution distance is used as the sorting algorithm, and the standardized calculation formula is modified to obtain a more comprehensive and differentiated evaluation result, providing an accurate process for signal concealment performance evaluation. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] Figure 1 Flowchart for low-acquisition signal covert performance evaluation.

[0058] Figure 2 Schematic diagram for the hierarchical selection of indicators.

[0059] Figure 3 This is the AHP weight diagram before improvement.

[0060] Figure 4 This is the improved AHP weight diagram.

[0061] Figure 5 It is a weight comparison chart of the entropy weight method before and after improvement.

[0062] Figure 6 This is a graph of the evaluation results of 11 signals under noise-free conditions.

[0063] Figure 7 These are some of the result graphs showing the increase and invariance of concealment performance of each signal under different signal-to-noise ratios.

[0064] Figure 8 These are some results showing the degradation of concealment performance of each signal under different signal-to-noise ratios. DETAILED DESCRIPTION

[0065] The present invention is further described below in conjunction with the accompanying drawings.

[0066] Figures 1 to 8 The best embodiment of the present invention is shown below in conjunction with the attached Figures 1 to 8 The present invention is further described.

[0067] Specific: such as Figure 1 As shown: A method for evaluating the concealment performance of a low intercept signal comprises the following steps:

[0068] Step 1) Eleven typical low intercept signals are selected as examples to demonstrate the evaluation method. The specific parameters of the signals are shown in Table 1, where FMCW signal is a typical chirp spread spectrum signal, costas is a classic frequency hopping signal, frank and P1-P4 codes are multi-phase coded signals, and T1-T4 codes are multi-time code signals.

[0069] Table 1 Signal parameters

[0070]

[0071] This example evaluates and ranks the concealment performance of signals under noise-free and noisy conditions. Under noisy conditions, 121 schemes with a signal-to-noise ratio in the range of 0-20dB in a Gaussian white noise channel are selected.

[0072] The specific type of input signal and modulation parameters are based on Figure 2 Classify the indicators and calculate the values ​​of each indicator.

[0073] Power spectrum exceeds value a 1 Defined as: signal power spectrum P I (f) Exceeding the mean of the noise power spectrum The ratio of the part to the power spectrum length. The index is defined as

[0074]

[0075] in, is the mean function, N l This indicator measures the ability of the signal power spectrum to hide in noise;

[0076] Power spectrum normalized deviation a 2 It is defined as: taking the sum of multiple peaks in the central frequency band of the signal power spectrum density as the benchmark, and comparing the difference with the white noise power spectrum benchmark, the normalized deviation is obtained:

[0077]

[0078] Among them, ζ can take a positive value, generally 20, P i (f) is the normalized power spectrum of white noise, P fmax (f) is the normalized power spectrum of the signal; the index deviation a 2 Measures the number and distribution of signal power spectrum peaks;

[0079] The instantaneous intra-pulse feature b is to normalize the signal amplitude. Extract its instantaneous amplitude, instantaneous frequency and instantaneous phase, and calculate the variance of each sequence as σ a , σ f and A monotonically decreasing function cos(arctan(σf)) is set to represent the concealment effectiveness. The index range is (0-1). The larger the value, the better the stealth effect. The index is determined as:

[0080]

[0081] Where N is the signal length, σ a , σ f and represent the variance of instantaneous amplitude, instantaneous frequency and instantaneous phase respectively;

[0082] The fractal dimension of the indicator CWD is c 1 =D B +D I , where D B With D I Represent the box dimension and information dimension respectively. The signal needs to be processed by CWD transformation and then combined with fractal dimension to represent the information. First, CWD image preprocessing is performed, the time-frequency two-dimensional image of the signal is grayed and binarized, and the global threshold algorithm is used to process it to obtain the denoised binary time-frequency image. The box dimension and information dimension are introduced to measure the complexity and irregularity of the grayscale image. The box dimension is calculated by placing the signal in a uniformly divided grid and finding the minimum number of grids required to cover the set. It is known that the sequence after signal processing is f(k) (k=1,2,…,N). Let the minimum number of grids covered be:

[0083]

[0084] Where δ is the length of the grid side, δ = 1 / N. Therefore, the box dimension can be defined as: D B =-lnN l / lnδ, where δ is the grid side length of each box, N l is the number of boxes needed to cover the spectrogram object.

[0085] When calculating the information dimension, we need to first reconstruct the signal y(i) = |f(i+1)-f(i)|, and then calculate the probability of the element falling into the box and the information entropy, as shown below:

[0086]

[0087] Among them, P i is the probability of an element falling into the box, the information entropy is I(δ), D I is the information dimension, Where P i is the probability of an element falling into the box, P i lgP i It is the calculation of information entropy.

[0088] Index of discrete instantaneous time-bandwidth product c 2 The formula is The discrete time-bandwidth product is defined as: is the time-frequency function of the denoised signal,

[0089] The bispectrum is the Fourier transform of the third-order cumulant of the signal and has twelve symmetric regions. To reduce the data calculation amount for convenient processing, the diagonal slice and diagonal integral of the bispectrum are studied, and the box dimension and information dimension are used to characterize its complexity and irregularity degree.

[0090] The bispectrum slice fractal dimension index is: d 1 = D I ×2 - (D B - 1); D I and D B are the box dimension and information dimension of the bispectrum slice respectively.

[0091] The bispectrum diagonal integral index is: d 2 = D I ×2 + 10×(D B - 1); D I and D B are the box dimension and information dimension of the bispectrum diagonal integral respectively.

[0092] Cyclic spectrum index where Y W (p, q) and V Noise (p, q) are the cyclic spectra of Gaussian white noise respectively.

[0093] Cepstrum index d 4 = D B + 10×D I , and the cepstrum of the signal is The periodicity of the signal cepstrum is also characterized by the box dimension and information dimension.

[0094] The numerical vector of various signal indexes finally obtained is Γ = [a 1 , a 2 , b, c 1 , c 2 , d 1 , d 2 , d 3 , d 4 .

[0095] Step 2) Create an analytic hierarchy process model. With the research of the concealment performance as the target layer, different domains are divided into the criterion layer, and the indexes of each domain class are the index layer, and there is no cross between each scheme layer. Calculate the subjective weight coefficients of the specific layer indexes under each signal.

[0096] Establish a comparison matrix Z: Authoritative experts compare the indicators of the same layer according to the three-scale method:

[0097]

[0098] Calculate the importance ranking index r i : n 1 is the number of indicators at this level, r i is the sum of the elements in the ith row of the matrix Z, and rmax = max{ri}, rmin = min{ri}, then construct the expert judgment matrix A k ,k=1,2,…,m 1 , m 1 is the total number of experts, and the elements of each judgment matrix are composed of:

[0099]

[0100] Among them, g is the base point comparison degree, which selects the two most important and least important indicators in this level compared with the previous level for comparison, with values ​​ranging from 1 to 9. Different experts can adopt the same comparison degree.

[0101]

[0102] Find the optimal transfer matrix C of each expert judgment matrix k , Calculate the quasi-optimal transfer matrix D k , Calculate the indicator weights obtained by the specific expert judgment matrix at this level The indicator calculation formula is:

[0103]

[0104] Introducing the cloud model multi-expert collaborative improvement weight design method: It is known that each judgment matrix has n 1 Indicators, by m 1 The weight vector obtained by the expert scoring is in Assume the weight vector after cloud processing is The calculation formula is: in:

[0105]

[0106] The weight vector obtained by the optimal transfer matrix and cloud model improvement includes the criterion layer w Crit =[ω A ,ω B ,ω C ,ω D ], solution layer The subjective weights of each index in the scheme layer relative to the highest target layer in the improved analytic hierarchy process are as follows:

[0107]

[0108] As Figure 3 shown in Figure 4 the figure, the index weights before and after the improved analytic hierarchy process are compared. In the traditional algorithm, 4 experts and 1 volunteer are selected to score the criterion layer and the index layer. Although it only focuses on the pairwise comparison of the same-level indexes, the mutual importance between two indexes often affects the third index, and it seriously depends on the scores of authoritative experts and is affected by the subjective preferences of different experts. The improved analytic hierarchy process has higher sensitivity to the parameter changes with larger weights, effectively emphasizes the importance of such indexes in evaluating the implicit performance, the changes of the index weights are more consistent, no longer rely on the subjective judgment of a single expert, achieves the balance of evaluation, and moreover, the variance of the weight distribution obtained by different experts' evaluations decreases and the stability improves.

[0109] Step 3), calculate the objective weight coefficient by using the improved entropy weight method.

[0110] Given m schemes and n indexes, then construct the index value matrix as D = [Γ 1 , Γ 2 , Γ 3 , …, Γ m T , and standardize the data to obtain the matrix It is divided into two categories: cost-type indexes and benefit-type indexes . For the former, the smaller the value, the better, and for the latter, the larger the value, the better. In order to avoid the situation where the negative ideal solutions are all 0 during the evaluation, the present invention improves the standardization formula on the basis of range standardization and performs the arctangent function scaling process on the minimum value, where γ can take a relatively large positive value:

[0111]

[0112] After standardization, perform normalization and calculate the proportion of the sum of all values of each index under this type of index: Calculate the information entropy of each index It should be noted that if p ij = 0, then it is stipulated that E j = 0.

[0113] Calculate the correlation coefficient matrix between indexes. The calculation formula for the Kendall correlation coefficient between two indexes x and y:

[0114]

[0115] ​where ρ ij It represents the correlation coefficient between the i-th indicator and the j-th indicator, C represents the number of pairs of elements with consistency in indicators x and y, and D represents the number of pairs of elements with inconsistency. The Kendall correlation coefficient does not require the data to follow a specific distribution and is a non-parametric measurement method.

[0116] Then quantify the degree of conflict among the indicators.

[0117] To find the weight of each indicator, first calculate the information redundancy: v j =1-E j , then calculate the objective weight as:

[0118]

[0119] Figure 5 is the weight comparison of the entropy weight method before and after improvement. Obviously, indicator a 1 and a 2 It is the representation of power spectrum data at different angles. After the improvement, the weight is reduced, indicating that there is a strong positive correlation between the two. 1 and d 2 There are two dimensionality reduction processes of the bispectrum, and there are obvious differences. The improved weights reflect the differences and are more in line with the data characteristics of the bispectrum. 3 and d 4 It is the most commonly used indicator in traditional low-interception signal evaluation. The improved objective weight has been increased, highlighting the role of classic effectiveness evaluation.

[0120] Step 4) Based on game theory, the subjective and objective weights are combined to be weighted as follows: Solve for the coefficients to get the comprehensive weights.

[0121] Use the idea of ​​Nash equilibrium to find the best weight coefficient so that the deviation between each vector is minimized. (K=ψ,φ), according to the curve extreme value theorem and the differential properties of the matrix, the optimal first-order derivative condition is: This gives the corresponding linear equations:

[0122]

[0123] Normalize the coefficients: The final combined weight is:

[0124]

[0125] Step 5) Use the distance between good and bad solutions as the sorting method, and calculate the weighted matrix based on the comprehensive weight vector and the normalized index value: X = Y·W *, determine the positive ideal solution and the negative ideal solution are: Calculate the distance to the positive ideal solution: The distance to the negative ideal solution is the same, and then the relative fit of each solution is calculated. The solutions are sorted according to the degree of fit to obtain the final evaluation results.

[0126] like Figure 6 As shown in the figure, the concealment performance of various noise-free signals is evaluated and ranked, and the calculation formulas of three types of distances, namely traditional Euclidean distance, Minkowski distance, and normalized projection distance, are compared. The vertical axis in the figure represents their scores, different colors represent different distance calculation formulas, and the values ​​on each column represent the ranking. The scores obtained by different distance calculation formulas in the figure are similar and the ranking is consistent, which indirectly proves the effectiveness of the proposed evaluation framework. The concealment performance ranking of typical noise-free signals is FMCW>P2>P1>frank>P4>P3>T2>costas>T3>T4>T1.

[0127] from Figure 7 and Figure 8 It can be seen that there is no obvious linear relationship between signal concealment performance and signal-to-noise ratio. This is the inherent characteristic of the modulated signal. The concealment performance of different types of signals varies greatly, and the concealment characteristics of signals of the same type show similar trends. Specifically, the evaluation results of frank code and FMCW increase because their indicators a 1 and a 2 As the signal-to-noise ratio increases, the evaluation results of T1-T4 code and costas signal decrease significantly because their index d 2 It decreases with the increase of signal-to-noise ratio; the various indicators of P1-P4 codes are very stable to noise, so the evaluation results are basically unchanged.

[0128] The present invention proposes a method for evaluating the concealment performance of low-intercept signals, which takes into account the three key steps of indicator selection, weight coefficient calculation and sorting algorithm, and realizes the measurement and comparison of the concealment performance of typical low-intercept signals. According to the three stages of signal interception, the present invention selects nine indicators from different fields, combines the optimal transfer matrix with the cloud model to effectively solve the problem of multi-expert collaboration and repeated consistency verification, introduces the correlation coefficient to improve the entropy weight method, examines the degree of conflict between indicators, and obtains more reasonable weights. It is proved that the concealment performance of the signal is an inherent attribute, which is constrained by the change of the signal's specific measurement parameters. Therefore, the present invention can compare the concealment characteristics between different signals, help the transmitter select the best signal waveform, and optimize the interception receiver to achieve efficient interception.

[0129] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A method for evaluating the concealment performance of a low intercept signal, characterized in that: The following steps are involved: Step 1: Select low interception signals, calculate various indicators in different domains for the selected signals according to the indicator layer diagram, construct a data matrix and calculate according to the improved standardized formula; Step 2: Create an improved AHP model, study the concealment performance as the target layer, divide different domains into the criterion layer, and the indicator layer contains a total of nine indicators in different domains, and then calculate the subjective weight coefficient; Step 3: Create an entropy weight method model, improve the standardized formula and introduce the correlation coefficient to calculate the objective weight of the indicator; Step 4: Balance the subjective and objective weights based on the combined weighting idea of ​​game theory to obtain a comprehensive indicator weight coefficient; Step 5: Use the distance between good and bad solutions as the sorting method, obtain the weighted matrix based on the comprehensive indicator weight coefficient and the normalized signal data, calculate the relative fit of each solution, and sort by fit to get the evaluation result.

2. A method for evaluating concealment performance of low interception signals according to claim 1, characterized in that: The index hierarchical diagram in step 1 includes an index layer, a criterion layer, a decision layer and a target layer, and the type and modulation parameters of the selected signal are input; the index vector of each signal is obtained as Γ=[a1,a2,b,c1,c2,d1,d2,d3,d4], and the data matrix D=[Γ1,Γ2,Γ3,…,Γ m ] T .

3. A method for evaluating concealment performance of low interception signals according to claim 2, characterized in that: The indicator layer specifically includes: The power spectrum exceedance value a1 is defined as: signal power spectrum P I (f) Exceeding the mean of the noise power spectrum The ratio of the part to the power spectrum length is defined as: in, is the mean function, N l This indicator measures the ability of the signal power spectrum to hide in noise; The power spectrum normalization deviation a2 is defined as: taking the sum of multiple peaks in the central frequency band of the signal power spectrum density as the benchmark, and comparing the difference with the white noise power spectrum benchmark, the normalized deviation is obtained: Among them, ζ can take a positive value, generally 20, P i (f) is the normalized power spectrum of white noise, P fmax (f) is the normalized power spectrum of the signal; the indicator deviation a2 measures the number and distribution of the peaks of the signal power spectrum; The instantaneous pulse feature b is to normalize the signal amplitude; extract its instantaneous amplitude, instantaneous frequency and instantaneous phase, and calculate the variance of each sequence as σ a , σ f and Assume a monotonically decreasing function cos(arctan(σ f )) represents the concealment effectiveness. The index range is (0-1). The larger the value, the better the stealth effect. The index is determined as: Where N is the signal length, σ a , σ f and Represent the variance of instantaneous amplitude, instantaneous frequency and instantaneous phase respectively; The fractal dimension of the indicator CWD is c1=D B +D I , where D B With D I represent the box dimension and information dimension respectively; The box dimension is defined as: D B =-lnN l / lnδ, where δ is the grid side length of each box, N l is the number of boxes needed to cover the frequency spectrum object; Information Dimension D I Defined as: Where P i is the probability of an element falling into the box, P i lgP i It is the calculation of information entropy; The formula for the discrete instantaneous bandwidth product c2 is: The discrete time-bandwidth product is defined as: in, is the time-frequency function of the denoised signal, To find the mean function, To find the variance function; and They are unit bandwidth and unit time respectively; The fractal dimension index of the bispectral slice is: d1 = D I ×2-(D B -1); D I With D B They are the box dimension and information dimension of the bispectral slice, respectively; The bispectral diagonal integral index is: d2 = D I ×2+10×(D B -1); D I With D B They are the box dimension and information dimension of the diagonal integral of the bispectrum, respectively; Cycle spectrum index where Y W (p,q) and V Noise (p,q) are the cyclic spectra of Gaussian white noise respectively; Cepstrum index d4 = D B +10×D I , the cepstrum of the signal is The box dimension and information dimension are used to characterize the periodicity of the signal cepstrum.

4. The method for evaluating the concealment performance of a low intercept signal according to claim 1, characterized in that: The step 2 constructs a hierarchical analysis method model, and the indicators of each layer are obtained by scoring by multiple experts. The weight vector obtained by improving the optimal transfer matrix and the cloud model includes the criterion layer w Crit =[ω A ,ω B ,ω C ,ω D ], solution layer and The subjective weight of each solution level indicator relative to the highest target level in the improved analytic hierarchy process is calculated as follows:

5. The method for evaluating the concealment performance of a low intercept signal according to claim 1, characterized in that: In step 3, an entropy weight method model is constructed, and a correlation coefficient is introduced to calculate the index conflict degree. The standardization formula is improved on the basis of range standardization, and the minimum value is scaled by an inverse tangent function, where γ takes a larger positive value: Calculate the correlation coefficient matrix between indicators; the Kendall correlation coefficient calculation formula between two indicators x and y is: Among them, ρ ij represents the correlation coefficient between the i-th indicator and the j-th indicator, C represents the number of element pairs with consistency in indicators x and y, and D represents the number of element pairs with inconsistency; the objective weight W of the entropy weight method is calculated φ .

6. The method for evaluating the concealment performance of a low intercept signal according to claim 1, characterized in that: Step 4 constructs comprehensive weights based on game theory The linear equations are as follows: Solve to get the comprehensive weight in and are the normalized linear coefficients.

7. The method for evaluating the concealment performance of a low intercept signal according to claim 1, characterized in that: The step 5 uses the superior and inferior solution distance as a sorting method, and first obtains the weighted sample matrix X=Y·W * , thus the positive and negative ideal solutions are as follows: Calculate the distance between each solution and the positive and negative ideal solutions and The final relative fit is: The solutions are sorted according to the degree of fit to obtain the final evaluation results.

8. A computer device / equipment / system comprising a memory, a processor and a computer program stored in the memory, characterized in that: The processor executes the computer program to implement the steps of the method according to any one of claims 1 to 7.

9. A computer-readable storage medium having a computer program / instruction stored thereon, characterized in that: When the computer program / instructions are executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

10. A computer program product comprising a computer program / instructions, characterized in that: When the computer program / instructions are executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

Images