A comprehensive evaluation method for communication anti-jamming performance

Abstract

The application relates to a comprehensive evaluation method for communication anti-interference performance and belongs to the technical field of communication. The application adopts various algorithms such as data standardization, grey correlation degree calculation and weight calculation, unifies indexes of different anti-interference systems or schemes to one dimension, and calculates the advantages and disadvantages. The application introduces data correlation characteristics into the comprehensive evaluation of the index system, can objectively evaluate the advantages and disadvantages of the comprehensive performance of different systems through a mathematical method, avoids the uncertainty of subjective scoring in the traditional method, and significantly improves the accuracy and reliability of the comprehensive performance evaluation.

Description

Technical Field

[0001] This invention belongs to the field of communication technology, and specifically relates to a comprehensive evaluation method for communication anti-interference performance. Background Technology

[0002] With the rapid development of wireless communication technology, spectrum resources are becoming increasingly scarce, especially in applications such as 5G, spread spectrum communication, frequency hopping communication, and narrow beam communication. The electromagnetic environment is becoming increasingly complex, with various types of interference signals frequently appearing. This has led to the rapid development of anti-interference communication technologies. Current anti-interference communication mainly focuses on spatial, time, frequency, and energy domains, employing different anti-interference schemes for different interference types to achieve optimal anti-interference performance. However, the time-varying nature of interference sources and patterns makes it difficult for an anti-interference device to achieve optimal performance under different interference patterns. Therefore, a comprehensive evaluation method for anti-interference communication devices is needed. This method should integrate multiple indicators based on the performance metrics achieved by the device under different interference patterns to scientifically and objectively evaluate the device's anti-interference performance and select the optimal solution.

[0003] Existing comprehensive evaluation methods for interference resistance performance are mostly subjective, where experts directly assign weights to different types of interference resistance indicators, and then calculate scores and rankings after weighting each indicator. Alternatively, a subjective evaluation method based on the Analytic Hierarchy Process (AHP) is used: AHP first decomposes the problem into multiple independent indicators; then, multiple experts compare all indicators pairwise to determine the relative importance of each indicator, thus providing quantitative estimates, and using mathematical methods to determine the weights of each indicator; finally, the performance ranking of multiple devices is calculated comprehensively. It is evident that subjective evaluation methods are limited by the subjective cognition or experience of experts, and the conclusions given by different experts may vary significantly, lacking universality.

[0004] The Entropy Weight Method (EW) is an objective evaluation method that determines the weight of an indicator by quantifying its information entropy. The smaller the entropy value, the greater the dispersion of the indicator, the more information it provides, and the greater its weight; conversely, the larger the entropy value, the smaller the weight. Since the entropy weight method relies entirely on the differences between similar indicators across multiple different systems, it is difficult to avoid situations where individual data anomalies can cause significant deviations in the calculation results. Therefore, its objective evaluation accuracy is greatly affected by data fluctuations.

[0005] Therefore, existing technologies have significant uncertainties when facing complex indicator systems with multiple types of interference and multi-dimensional evaluation. There is an urgent need to propose a new objective evaluation method for the comprehensive performance of communication anti-interference to solve the above problems. Summary of the Invention

[0006] (a) Technical problems to be solved

[0007] The technical problem to be solved by this invention is how to provide a comprehensive evaluation method for communication anti-interference performance, so as to solve the problem of large uncertainty in the existing technology in the face of complex index system with multiple types of interference and multi-dimensional evaluation.

[0008] (II) Technical Solution

[0009] To address the aforementioned technical problems, this invention proposes a comprehensive evaluation method for communication anti-interference performance, which includes the following steps:

[0010] S1. Determine all anti-interference index types based on the types of interference sources. Input N performance indices for M systems or schemes to form an M×N matrix D. Extract the optimal and least desirable sequences from matrix D and combine them with matrix D to form the optimal sequence matrix and the least desirable sequence matrix, respectively. Perform normalized operations on the optimal and least desirable sequence matrices to obtain the standardized matrices. ;

[0011] S2, Grey Relational Degree: Using a standardized matrix Calculate the grey relational coefficient matrices of the input M×N matrix and the optimal and least ordered sequences, respectively. ;

[0012] S3, Based on Standardized Matrix Calculate the normalized weights of N indicators ;

[0013] S4, Based on the grey relational coefficient matrix and normalized weights Obtain the grey relational degree between each system or scheme and the optimal and worst sequences. ;

[0014] S5. Optimize the evaluation model: For Normalization yields The evaluation model is obtained by weighting the components, constructing the cost function using the least squares method, and finding the extreme values ​​under constraints using the Lagrange multiplier method, then adjusting the... The weights are adjusted to make the model as close as possible to the optimal sequence and as far away as possible from the worst sequence;

[0015] S6. Based on the weights calculated in S5, the final evaluation model is obtained, and the final comparison results of the anti-interference performance of each system or scheme are obtained.

[0016] (III) Beneficial Effects

[0017] This invention proposes a comprehensive evaluation method for communication anti-interference performance. The beneficial effects of this invention are as follows:

[0018] (1) Construct an objective evaluation model for common communication anti-interference index system

[0019] This invention incorporates data correlation characteristics into the comprehensive evaluation of the indicator system, enabling objective evaluation of the overall performance of different systems through mathematical methods. This avoids the uncertainty of relying on subjective scoring in traditional methods and significantly improves the accuracy and reliability of the comprehensive performance evaluation.

[0020] (2) It has adaptive feature extraction and modeling capabilities.

[0021] This invention employs principal component analysis (PCA) to adaptively eliminate the correlation between various indicators, forming mutually independent principal components. Based on the orthogonal nature of these principal components, calculating one factor does not introduce the influence of other factors, thus making the adjustment of the weights of various indicators more reasonable and accurate, and resulting in a superior evaluation model.

[0022] (3) It does not rely on the distribution assumptions of the sample and has the advantage of small sample size analysis.

[0023] This invention employs the grey relational analysis algorithm to calculate the correlation between each system index and the optimal and least optimal sequences. It requires a small sample size, making it suitable for small sample size data analysis. Furthermore, it does not rely on assumptions such as the normal distribution of the data and has a significant effect on atypical data.

[0024] Through the above-mentioned innovative strategies and technologies, the method described in this invention can effectively overcome the limitations of traditional subjective evaluation methods for comprehensive performance, which are restricted by experience or cognition, and the shortcomings of objective evaluation methods in detecting abnormal data. It has strong applicability, accuracy and application value in the comprehensive performance evaluation of communication anti-interference. Attached Figure Description

[0025] Figure 1 This is a flowchart of the method of the present invention. Detailed Implementation

[0026] To make the objectives, contents, and advantages of the present invention clearer, the specific embodiments of the present invention will be described in further detail below with reference to the accompanying drawings and examples.

[0027] This invention belongs to the field of wireless anti-interference communication technology, specifically involving an improved calculation method based on grey relational analysis to achieve a comprehensive and objective evaluation of anti-interference performance. It can be applied to the selection of wireless anti-interference communication design schemes and the performance evaluation of anti-interference equipment.

[0028] The purpose of this invention is to propose an objective evaluation method for the comprehensive anti-interference performance of communication systems. This method, even with limited simulation or measurement data, unifies indicators from different dimensions within a single framework to determine the merits of different communication systems or solutions. Evaluators can make optimal decisions by comprehensively weighing the anti-interference costs of different system solutions.

[0029] Specifically, this invention employs multiple algorithms such as data standardization, grey relational analysis, and weight calculation to unify the indicators of different anti-interference systems or schemes into a single dimension and calculate their merits. This invention includes the following steps:

[0030] S1. Determine all types of anti-interference indicators (N types) based on the types of interference sources. Input the N performance indicators of M systems or schemes to form an M×N matrix D. Extract the optimal and worst (inferior) sequences from matrix D, and combine them with matrix D to form the optimal sequence matrix and the worst sequence matrix, respectively. Perform normalized operations on the optimal sequence matrix and the worst sequence matrix to obtain the standardized matrix. .

[0031] S2, Grey Relational Degree: Using a standardized matrix Calculate the grey relational coefficient matrices of the input M×N matrix and the optimal and least ordered sequences, respectively. ;

[0032] S3, Based on Standardized Matrix Calculate the normalized weights of N indicators ;

[0033] S4, Based on the grey relational coefficient matrix and normalized weights Obtain the grey relational degree between each system or scheme and the optimal and worst sequences. .

[0034] S5. Optimize the evaluation model: For Normalization yields The evaluation model is obtained by weighting the components, constructing the cost function using the least squares method, and finding the extreme values ​​under constraints using the Lagrange multiplier method, then adjusting the... The weights are adjusted to make the model as close as possible to the optimal sequence and as far away as possible from the worst sequence;

[0035] S6. Based on the weights calculated in S5, the final evaluation model is obtained, and the final comparison results of the anti-interference performance of each system or scheme are obtained.

[0036] Furthermore, the specific steps of the objective evaluation method for the overall anti-interference performance of communication are as follows:

[0037] Step S1, data acquisition and preprocessing, specifically includes:

[0038] S11. Collect N performance indicators of M systems or schemes under spatial anti-interference, frequency anti-interference, and time-domain anti-interference conditions, and form an M×N matrix D.

[0039] S12. Determine the reference sequence Extract the optimal sequence of N performance metrics. and the most recent sequence In this process, the optimal value for each performance index is formed into an optimal sequence, and the worst value is formed into a worst sequence.

[0040] S13, will As reference sequences, they are combined with matrix D to form an (M+1)×N matrix, which is the optimal sequence matrix. and least significant sequence matrix .

[0041] S14, to After standardization, the (M+1)×N matrix to be standardized is represented as:

[0042]

[0043] in, This is the reference sequence.

[0044] Standardize X to a matrix that follows a normal distribution, i.e.:

[0045]

[0046] in:

[0047]

[0048]

[0049] Finally, the standardized matrix is ​​obtained. .

[0050] Step S2: Calculate separately The grey relational coefficient matrix is ​​defined as follows:

[0051]

[0052] in:

[0053]

[0054] For calculation Calculations were performed to obtain .

[0055] Step S3: Based on the standardized matrix respectively calculate Weights of N internal indicators

[0056] S31, calculate separately The autocovariance matrix is ​​expressed as:

[0057]

[0058] because All data are standardized, and the mean of their column vectors has been reduced to zero. Therefore:

[0059]

[0060] Calculated .

[0061] S32. Solve the matrices respectively. Eigenvalues ​​and eigenvectors:

[0062] For an N-order square matrix, its eigenvalues, from largest to smallest, are represented as follows: The corresponding vector representation is Take the first d eigenvalues, where d is the minimum value that satisfies the following equation:

[0063]

[0064]

[0065] The vector that makes up the eigenvalues:

[0066]

[0067] Take the corresponding feature vector (Column vectors) form an N×d matrix:

[0068]

[0069] S33, calculate separately The weights of each internal indicator were assigned and normalized.

[0070] Calculate the weights:

[0071]

[0072] The weights are normalized, and the normalized weights are:

[0073]

[0074] in:

[0075]

[0076] in, This is the floor function.

[0077] Finally, we obtained the corresponding to Normalized weights .

[0078] Step S4: Calculate separately Grey relational degree :

[0079]

[0080] It is a 1×M vector.

[0081] Step S5: Model optimization, making the index as close as possible to the optimal sequence and as far away from the suboptimal sequence.

[0082] S51, to Normalization:

[0083]

[0084] For a 1×M vector:

[0085]

[0086] S52, to The final evaluation model is obtained by weighting the components separately:

[0087]

[0088] in: As the weight, satisfying the condition:

[0089]

[0090] S53. Solving for the optimal weight values:

[0091] Construct the cost function using the least squares method:

[0092]

[0093] Using the Lagrange multiplier method to find conditional extrema, establish the Lagrange function:

[0094]

[0095] To each Taking the first-order partial derivative and setting it equal to 0, we get:

[0096]

[0097] Solving the system of equations yields the following:

[0098]

[0099] Step S6: Ranking by merit

[0100] Final evaluation model , Given a 1×M array, its elements The scores represent the final evaluation scores of systems (or schemes) 1, 2, ..., M, respectively. Arranging these scores from highest to lowest gives the final ranking of the anti-interference performance of each system (or scheme).

[0101] This invention selects the system or scheme with the best anti-interference performance by determining which system or scheme's indicators are best associated with the optimal sequence and worst associated with the worst sequence.

[0102] Example 1:

[0103] Implementation steps and technical details:

[0104] Step 1: Data preparation, including:

[0105] (1) Extract indicators from multiple actual communication systems or multiple simulation schemes. Taking eight similar frequency hopping anti-interference systems as an example, their anti-interference indicators are as follows:

[0106] Spatial interference immunity: Interference immunity tolerance was measured at 10°, 20°, and 30° in both the signal direction and the included angle. Since the number of array elements and the array arrangement were the same, the 30° included angle reached the maximum value of the spatial interference immunity tolerance.

[0107] Frequency domain interference immunity: Under the condition that spatial domain interference immunity is ineffective (the angle between the signal and interference directions is 0°), assuming that some operating frequency bands are subjected to full-time suppression interference, measure the maximum acceptable proportion of the interference-resistant frequency band for each system. Under the conditions of 40% and 30% frequency bands subjected to suppression interference, measure the achievable limit sensitivity of each system.

[0108] Temporal interference immunity: Under the condition that spatial interference immunity is ineffective (the angle between the signal and interference directions is 0°), assuming that the system is subjected to full-band suppression interference for part of the operating time, the maximum proportion of interference immunity time that each system can withstand is measured. Under the conditions of 40% and 30% time subjected to suppression interference, the limit sensitivity that each system can achieve is measured.

[0109] Table 1. Example data for Example 1.

[0110]

[0111] (2) Determine the optimal sequence and the least optimal sequence. Among them, the anti-interference tolerance, anti-interference bandwidth and anti-interference time index are high-optimal (the higher the value, the better), and the anti-interference sensitivity index is low-optimal (the lower the value, the better).

[0112]

[0113] (3) Data standardization:

[0114]

[0115]

[0116] Step 2: Calculate separately The coefficients of the grey relational matrix:

[0117]

[0118]

[0119] Step 3: Calculate separately Normalized weights for each indicator:

[0120]

[0121]

[0122] Step 4: Calculate separately Grey relational degree :

[0123]

[0124]

[0125] Step 5: Model Optimization and Ranking:

[0126] (1) Construct the cost function and solve for the weights of the optimal and least-order grey relational degrees:

[0127]

[0128] (2) The final evaluation model is:

[0129]

[0130] (3) Vector The final evaluation scores for systems 1 through 8 are listed in descending order, and their relative merits are as follows:

[0131]

[0132] The beneficial effects of this invention are as follows:

[0133] (1) Construct an objective evaluation model for common communication anti-interference index system

[0134] This invention incorporates data correlation characteristics into the comprehensive evaluation of the indicator system, enabling objective evaluation of the overall performance of different systems through mathematical methods. This avoids the uncertainty of relying on subjective scoring in traditional methods and significantly improves the accuracy and reliability of the comprehensive performance evaluation.

[0135] (2) It has adaptive feature extraction and modeling capabilities.

[0136] This invention employs principal component analysis (PCA) to adaptively eliminate the correlation between various indicators, forming mutually independent principal components. Based on the orthogonal nature of these principal components, calculating one factor does not introduce the influence of other factors, thus making the adjustment of the weights of various indicators more reasonable and accurate, and resulting in a superior evaluation model.

[0137] (3) It does not rely on the distribution assumptions of the sample and has the advantage of small sample size analysis.

[0138] This invention employs the grey relational analysis algorithm to calculate the correlation between each system index and the optimal and least optimal sequences. It requires a small sample size, making it suitable for small sample size data analysis. Furthermore, it does not rely on assumptions such as the normal distribution of the data and has a significant effect on atypical data.

[0139] Through the above-mentioned innovative strategies and technologies, the method described in this invention can effectively overcome the limitations of traditional subjective evaluation methods for comprehensive performance, which are restricted by experience or cognition, and the shortcomings of objective evaluation methods in detecting abnormal data. It has strong applicability, accuracy and application value in the comprehensive performance evaluation of communication anti-interference.

[0140] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the technical principles of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A method for comprehensive evaluation of communication anti-jamming performance, characterized in that, The method comprises the following steps: S1. Determine all anti-interference index types based on the type of interference source, and input... M A system or solution N These performance indicators form M × N matrix D From the matrix D Extract the optimal and least desirable sequences from the matrix, and then compare them with the matrix. D The optimal sequence matrix and the least ordered sequence matrix are formed by combining them; then, the optimal sequence matrix and the least ordered sequence matrix are standardized according to a normal distribution to obtain the standardized matrix. ; S2, Grey correlation degree: use the standardized matrix Calculate the input M × N Matrix of grey correlation coefficient of the matrix with the optimal sequence and the worst sequence ; S3, based on the standardized matrix computing N a normalization weight for the indicator ; S4, based on the grey correlation coefficient matrix and the normalized weight get the grey correlation degree of each system or scheme with the optimal sequence and the worst sequence ; S5, optimizing evaluation model: to normalization and weighted to get the evaluation model, using the least square method to construct the cost function, according to the Lagrange multiplier method to find the extreme value under the condition of restriction, adjust the weight of, make the model as close to the optimal sequence as possible and far away from the last sequence; S6, obtaining a final evaluation model according to the weight calculated in S5, and obtaining a final comparison result of anti-interference performance of each system or scheme; Wherein, The S1 comprises the following steps: S11, collect M The system or scheme forms N performance indexes under space anti-interference, frequency domain anti-interference and time domain anti-interference M × N Matrix D ; S12, determining a reference sequence : extracting N an optimal sequence of performance indicators and a worst sequence ; S13, to as the reference sequence respectively with matrix D composition( M +1)× N matrix, namely the optimal sequence matrix and the last sequence matrix ; S14, to standardized, the (S14) to be standardized M +1) x N matrix representation is: wherein, is a reference sequence; The X standardized to a matrix that conforms to a normal distribution, i.e.: Wherein: ； ； ； ； ； resulting in a normalized matrix .

2. The method of claim 1, wherein the communication anti-jamming performance is evaluated by: In the S12, the optimal value of each performance index forms an optimal sequence, and the worst value forms a last sequence.

3. The comprehensive evaluation method for communication anti-interference performance as described in claim 1, characterized in that, The S2 comprises: The grey correlation coefficient matrix is defined as: Wherein: ； ； ； ； ； ； Respectively for The calculation is performed to obtain .

4. The comprehensive evaluation method for communication anti-interference performance as described in claim 3, characterized in that, The S3 comprises: S31, respectively calculating the autocovariance matrix of the signal, denoted as: Since are standardized data, whose column vectors have been zeroed, so that: calculated ; S32, solve the eigenvalues and eigenvectors of matrix respectively: For an N x N matrix, the eigenvalues are ordered from largest to smallest as The corresponding eigenvectors are The first d eigenvalues are taken as d The minimum value of the following equation is found: The vector composed of the characteristic values: take the corresponding column vector compositions N x d matrix: S33, respectively calculating the weight of each index, and normalizing Calculate the weight: The normalized weight is: Wherein: wherein floor function; The final normalized weights corresponding to are obtained.​ 5. The method of claim 4, wherein the communication anti-jamming performance is evaluated by: In the S32, .

6. The method of claim 5, wherein the communication anti-jamming performance is evaluated by: The S4 comprises: is 1 x M Vector.

7. The method of claim 6, wherein the communication anti-jamming performance is evaluated by: The S5 comprises: S51, to Normalization: For 1 x M Vector: S52, to Respectively weighted to obtain final evaluation model: wherein: are weights, satisfying the condition: S53, solving the optimal weight value: A cost function is constructed by the least square method: According to the Lagrange multiplier method, a Lagrange function is established: Respectively to The first-order partial derivative is obtained and set equal to 0, and the following is obtained: The equation set can be solved to obtain: 。 8. The method of claim 7, wherein the communication anti-jamming performance is evaluated by: The S6 comprises: a final evaluation model , is 1 x M array, elements represent the evaluation final score of the system or scheme 1, 2, …, M in turn, arranged in descending order of score, that is, the final ranking of the anti-interference performance of each system or scheme.

9. The method of claim 1, wherein, The anti-interference indexes include: Spatial anti-interference: the anti-interference tolerance in 10°, 20°, 30° of signal direction and angle direction is measured respectively; Frequency domain anti-interference: in the case of invalid spatial anti-interference, it is assumed that part of the working frequency band is subjected to full-time suppression type interference, the maximum anti-interference frequency band proportion of each system is measured, and in the case of 40%, 30% frequency band subjected to suppression type interference, the limit sensitivity that can be reached by each system is measured; Time domain anti-interference: in the case of invalid spatial anti-interference, it is assumed that part of the working time is subjected to full-frequency suppression type interference, the maximum anti-interference time proportion of each system is measured, and in the case of 40%, 30% time subjected to suppression type interference, the limit sensitivity that can be reached by each system is measured.

Images