A radar narrowband interference recognition method based on envelope histogram distribution

Abstract

The present invention relates to a radar narrowband interference identification method based on envelope histogram distribution. Aiming at the shortcomings of traditional radar interference type confirmation and anti-interference measure selection, which rely more on manual judgment and setting and have ambiguity and subjectivity, a method for accurately classifying four typical narrowband interferences, namely narrow pulse interference, suppression interference, deceptive interference and convolution interference, is proposed. The specific steps include: echo amplitude envelope construction, equal amplitude histogram distribution statistics, adjacent window refinement compensation and adaptive classification and recognition. The present invention effectively enhances the automatic selection capability of anti-interference measures and improves the combat effectiveness of radar in complex electromagnetic environments.

Description

Technical Field

[0001] The invention relates to an interference identification method, which is mainly applied to the field of radar interference signal classification and identification. Background Art

[0002] Realizing accurate and real-time perception of spatial interference situations in complex electromagnetic environments is an important prerequisite for optimizing the selection of countermeasure strategies and refining and layering interference suppression in the process of radar's game with electronic warfare equipment. With the continuous development of electronic warfare equipment, typical interference styles such as narrow pulse interference, suppression interference, deceptive interference and convolution have been proposed one after another, posing new challenges to modern radar's ability to perceive battlefield interference environments.

[0003] The document "Research and Implementation of Radar Active Interference Signal Identification Method" (Beijing Institute of Technology) uses digital channelization technology to obtain the time-frequency domain characteristic parameters and pulse description word parameters of the interference signal for common radar active interference signals, and designs DSP-based radar common active interference signal identification software based on this; the document "Research and Implementation of Radar Interference Signal Feature Identification Technology" (National University of Defense Technology) conducts research on radar interference signal feature recognition technology, including research on radar interference signal modulation type recognition algorithms, and engineering implementation technology of broadband radar interference signal digital receivers; the patent "A Color Image Segmentation Method Based on Adaptive Hierarchical Histogram" (CN201811517182.4) provides a color image segmentation method based on an adaptive hierarchical histogram, which has better segmentation efficiency and can obtain better segmentation quality; the patent "Image Processing Method and System Based on Integral Histogram" (CN200810167131.3) provides an image processing method based on integral histogram. Summary of the invention

[0004] In order to overcome the shortcomings of traditional radars that they rely more on manual judgment and setting in interference identification and the selection of anti-interference measures, which have strong ambiguity and subjectivity, a radar narrowband interference identification method based on envelope histogram distribution is proposed by constructing equal-amplitude histograms and analyzing the distribution characteristics of the compensated histograms.

[0005] In order to achieve the purpose of the present invention, a radar narrowband interference identification method based on envelope histogram distribution includes the following steps, and the implementation process is as follows:

[0006] S1: Echo amplitude envelope construction. The radar original zero intermediate frequency echo data x is modulo and then logarithmically processed:

[0007] y=A×logB(|x|) (8)

[0008] Wherein, y represents the logarithmic result of the echo amplitude envelope, A represents the amplitude coefficient, logB(·) represents the logarithm with B as the base, and |·| represents the modulus value of the vector element.

[0009] S2: Equal amplitude histogram distribution statistics. Solve the maximum value y of the logarithmic result y of the echo amplitude envelope max and the minimum value y min . Set the equal amplitude step size of the equal amplitude histogram to T unit , divide the histogram into N unit Grids:

[0010] N unit =floor((y max -y min ) / T unit )+1 (9)

[0011] Where floor(·) means rounding towards negative infinity.

[0012] With y min is the starting point of the histogram grid, T unit is the histogram grid step size, traverse each element in the logarithmic result y of the echo amplitude envelope, and map it to the histogram grid:

[0013] n=floor((y m -y min ) / T unit )+1,m∈[1,M] (10)

[0014] S(n)=S(n)+1,n∈[1,N unit ] (11)

[0015] Wherein, M represents the total number of elements in the logarithmic result y of the echo amplitude envelope, n represents the equal amplitude histogram grid index, and S(n) represents the number of elements in the nth histogram grid.

[0016] S3: Refined compensation of adjacent windows. Perform backward compensation on each histogram grid to obtain the compensation result:

[0017] Z(i)=S(i)+S(i+1),i∈[1,N unit -1] (12)

[0018] Among them, Z(i) represents the histogram adjacent window refinement compensation result.

[0019] S4: Adaptive classification and recognition. Set the histogram detection threshold T jam :

[0020] T jam =mean(y norm) + 3×std(y norm ) (13)

[0021] where y norm represents the result of modulus logarithmic conversion of the echo when the radar is not interfered, mean(·) represents calculating the mean value, and std(·) represents calculating the variance. The histogram bin index N jam corresponding to the detection threshold T Tj is:

[0022] N Tj = floor((T jam - y min ) / T unit ) + 1 (14)

[0023] Classify and identify the compensated data Z(i) for histogram index i ∈ [N Tj + 1, N unit - 1]:

[0024] S41: Traverse all bins. If there exists Z(i) > M / 2, it is false target interference; otherwise, jump to S42;

[0025] S42: Count the number of bins where Z(i) < 2 as N min2 and the sum of all elements in all bins as N sum = sum(Z(i)), where sum(·) represents summing. If N min2 > (N unit - N Tj ) / 2 and N sum < M / 2, it is narrow pulse interference; otherwise, jump to S43;

[0026] S43: Calculate the kurtosis K z of the compensated data Z(i). If K z is less than the set threshold T z , it is noise-like interference; otherwise, it is convolution-like interference.

[0027] Advantages of the present invention: By adopting a radar narrowband interference recognition method based on envelope histogram distribution, accurately classify four typical narrowband interferences, namely narrow pulse interference, suppression interference, deception interference, and convolution-like interference, effectively enhance the automatic selection ability of anti-interference measures, improve the combat effectiveness of the radar in complex electromagnetic environments, and have the advantages of small computational amount, fast recognition speed, high accuracy rate, simple process, and convenient engineering implementation. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] Figure 1The flowchart of a radar narrowband interference identification method based on envelope histogram distribution is shown in FIG. S1 to S4 in the figure correspond to the S1 to S4 processes described in the invention content.

[0029] Figure 2 for Figure 1 The specific implementation method of each step in . Where y represents the logarithmic result of the echo amplitude envelope, S(n) represents the number of elements in the nth histogram grid, and Z(i) represents the refined compensation result of the histogram adjacent window. DETAILED DESCRIPTION

[0030] The implementation process of the present invention is as follows Figure 1 As shown, the specific process is described as follows:

[0031] S1: Echo amplitude envelope construction. The radar echo envelope characterizes the energy distribution characteristics of the radar received echo signal. In order to obtain the radar echo envelope and reduce the dynamic range of the echo, the radar original zero intermediate frequency echo data x is modulo and then logarithmically processed:

[0032] y=A×logB(|x|) (15)

[0033] Wherein, y represents the logarithmic result of the echo amplitude envelope, A represents the amplitude coefficient, logB(·) represents the logarithm with B as the base, and |·| represents the modulus value of the vector element.

[0034] S2: Equal amplitude histogram distribution statistics. Envelope histogram distribution is a typical feature of active interference, which can reflect the difference in amplitude distribution between different types of interference signals. Based on this, the maximum value y of the logarithmic result y of the echo amplitude envelope is solved. max and the minimum value y min , with the maximum value, minimum value and the equal amplitude step size T of the equal amplitude histogram unit Map the data. First, divide the histogram into N unit Grids:

[0035] N unit =floor((y max -y min ) / T unit )+1 (16)

[0036] Where floor(·) means rounding towards negative infinity.

[0037] With y min is the starting point of the histogram grid, T unit is the histogram grid step size, traverse each element in the logarithmic result y of the echo amplitude envelope, and map it to the histogram grid:

[0038] n=floor((ym -y min ) / T unit )+1,m∈[1,M] (17)

[0039] S(n)=S(n)+1,n∈[1,N unit ] (18)

[0040] Wherein, M represents the total number of elements in the logarithmic result y of the echo amplitude envelope, n represents the equal amplitude histogram grid index, and S(n) represents the number of elements in the nth histogram grid.

[0041] S3: Refined compensation of adjacent windows. To avoid the amplitude of weak jitter being mistakenly divided into two adjacent histogram grids, resulting in insufficient number of elements in the same amplitude interval in a single grid, making classification and recognition unstable, backward compensation is performed on each histogram grid to obtain the compensation result:

[0042] Z(i)=S(i)+S(i+1),i∈[1,N unit -1] (19)

[0043] Among them, Z(i) represents the histogram adjacent window refinement compensation result.

[0044] S4: Adaptive classification and identification. Active interference is a one-way transmission. At the same time, electronic warfare equipment usually has strong interference energy in order to suppress radar detection of real targets. In order to avoid the histogram classification and identification parameters being affected by the radar base noise, the histogram detection threshold T is set. jam :

[0045] T jam =mean(y norm )+3×std(y norm ) (20)

[0046] Among them, y norm It represents the logarithmic result of echo modulus when the radar is not disturbed, mean(·) represents the mean value, and std(·) represents the variance. jam The corresponding histogram grid index N Tj :

[0047] N Tj =floor((T jam -y min ) / T unit )+1 (21)

[0048] For histogram index i∈[N Tj +1,N unit -1] to classify and identify the compensated data Z(i):

[0049] S41: The basic feature of false target interference is "continuous rectangular envelope after modulo operation". Traverse all cells. If there exists Z(i) > M / 2, it is false target interference; otherwise, jump to S42;

[0050] S42: The basic feature of narrow pulse interference is "strong and short pulse envelope after modulo operation". Count the number of cells N where Z(i) < 2 in the cells min2 and the sum N of all elements in the cells sum = sum(Z(i)), where sum(·) represents summation. If N min2 > (N unit - N Tj ) / 2 and N sum < M / 2, it is narrow pulse interference; otherwise, jump to S43;

[0051] S43: The histogram of noise-like interference is a normal distribution envelope, and the histogram of convolution-like interference is a random distribution. Calculate the kurtosis K of the compensated data Z(i) z , if K z is less than the set threshold T z , it is noise-like interference; otherwise, it is convolution-like interference.

Claims

1. A radar narrowband interference identification method based on envelope histogram distribution, Features: S1: Echo amplitude envelope construction; the radar original zero intermediate frequency echo data x is modulo and then logarithmized: y=A×log B(|x|) (1) Where x represents the original zero intermediate frequency echo data of the radar, y represents the logarithmic result of the echo amplitude envelope, A represents the amplitude coefficient, log B(·) represents the logarithm with B as the base, and |·| represents the modulus value of the vector element; S2: Equal amplitude histogram distribution statistics; solve the maximum value y of the logarithmic result y of the echo amplitude envelope max and the minimum value y min ; Set the equal amplitude step size of the equal amplitude histogram to T unit , divide the histogram into N unit Grids: N unit =floor((y max -y min ) / T unit )+1 (2) Among them, floor(·) means rounding towards negative infinity; With y min is the starting point of the histogram grid, T unit is the histogram grid step size, traverse each element in the logarithmic result y of the echo amplitude envelope, and map it to the histogram grid: n=floor((y m -y min ) / T unit )+1,m∈[1,M] (3) S(n)=S(n)+1,n∈[1,N unit ] (4) Wherein, M represents the total number of elements in the logarithmic result y of the echo amplitude envelope, n represents the grid index of the equal amplitude histogram, and S(n) represents the number of elements in the nth histogram grid; S3: Refine compensation of adjacent windows; perform backward compensation on each histogram grid to obtain the compensation result: Z(i)=S(i)+S(i+1),i∈[1,N unit -1] (5) Among them, Z(i) represents the refined compensation result of the histogram adjacent window, and S(i) represents the number of elements in the i-th histogram grid; S4: Adaptive classification and recognition; setting histogram detection threshold T jam : T jam =mean(y norm )+3×std(y norm ) (6) Among them, y norm represents the logarithmic result of echo modulus when the radar is not disturbed, mean(·) represents the mean value, and std(·) represents the variance; and the detection threshold T jam The corresponding histogram grid index N Tj : N Tj =floor((T jam -y min ) / T unit )+1 (7) For histogram index i∈[N Tj +1,N unit -1] to classify and identify the compensated data Z(i): S41: traverse all grids, if there is Z(i)>M / 2, it is a false target interference; otherwise jump to S42; S42: Count the number of grids N where Z(i) < 2 in the grid min2 , the sum of all elements in the grid N sum =sum(Z(i)), where sum(·) represents sum; if N min2 >(N unit -N Tj ) / 2 and N sum <M / 2, it is narrow pulse interference; otherwise jump to S43; S43: Calculate the kurtosis K of the compensated data Z(i) z , if K z Less than the set threshold T z , it is noise interference; otherwise it is convolution interference.

Images