A histogram-based pulse repetition interval estimation method, system, and storage medium

Abstract

The application provides a histogram-based pulse repetition interval estimation method, system and storage medium, which comprises the following steps: classifying radar reconnaissance signals according to carrier frequencies to obtain a data set of each carrier frequency; drawing a histogram of the data set and generating a corresponding statistical list; based on the statistical list, applying a PRI initial value selection method to generate a second PRI typical value; applying a boundary determination method to generate a second left boundary and a second right boundary corresponding to the second PRI typical value, and storing the positions of other PRI values to be processed in the statistical list in a second other vector; sorting the second other vector from small to large, and performing a loop processing until the second other vector is zero; and determining a final sorted PRI typical value and its left boundary and right boundary according to the loop processing result. The application does not require complex parameter selection, greatly reduces the workload of operators and professional requirements, and improves the sorting accuracy.

Description

Technical Field

[0001] This invention relates to the field of counter-reconnaissance technology, specifically to a pulse repetition interval estimation method, system, and storage medium based on histograms. Background Technology

[0002] Radar signal sorting is the process of separating pulses belonging to different radar radiation sources from a dense stream of intercepted radar pulses (i.e., full pulses, each pulse represented by a pulse descriptor word PDW). It is the foundation of radar electronic countermeasures intelligence analysis. Only by sorting the randomly overlapping pulse stream into individual pulse sequences of each radar can the characteristic parameters of the radar be accurately measured and analyzed in detail, thereby determining the functional purpose, platform type and / or threat level of these radars, and accurately jamming enemy threat radiation sources.

[0003] In current electronic countermeasures intelligence analysis, pre-sorting methods are mainly based on three dimensions of the signal: carrier frequency, angle, and pulse width. The space composed of the signal's characteristic parameters is divided into multiple sorting subspaces, and the pulse sets to be sorted are projected onto these subspaces to dilute the pulse stream. Traditional pre-sorting methods are mainly based on algorithms for deinterleaving the pulse repetition interval (PRI). Their fundamental basis is the inherent PRI characteristics of radar, and various estimation algorithms are all based on an understanding of these PRI characteristics, attempting to "match" them as closely as possible. Currently commonly used pre-sorting algorithms include sequence search, traditional statistical histograms, histogram of differences (CDIF), histogram of order differences (SDIF), and PRI transform methods, among others. The detection performance of different algorithms needs improvement in complex signal environments such as pulse loss, pulse jitter, and pulse unevenness. Common problems include: 1) difficulty in extracting the correct PRI feature values ​​and corresponding tolerances when a large number of signal patterns have adjacent and overlapping PRI tolerances; 2) long processing time, excessive computational load, and low efficiency. Summary of the Invention

[0004] In view of this, the present invention proposes a pulse repetition interval estimation method, system and storage medium based on histogram. Through the analysis of parameter estimation results under multiple sets of classification data, it is proved that the method is simple, practical and applicable to complex signal environments.

[0005] According to one aspect of the present invention, a histogram-based method for estimating pulse repetition intervals is provided, comprising:

[0006] S101: Classify radar reconnaissance signals according to carrier frequency to obtain a dataset for each carrier frequency;

[0007] S102: Draw a histogram based on the data set and generate a statistical list corresponding to the histogram;

[0008] S103: Based on the statistical list, the second typical PRI value is generated by applying the PRI initial value selection method;

[0009] S104: Apply the boundary determination method to generate the second left boundary and the second right boundary corresponding to the second typical PRI value, and store the positions of other PRI values ​​to be processed in the statistical list in the second other vector;

[0010] S105: Sort the second other vector in ascending order, and process it in a loop until the second other vector is zero;

[0011] The loop processing involves repeatedly executing the PRI initial value selection method and the boundary determination method, and updating the typical PRI value and its left and right boundaries.

[0012] S106: Based on the results of the cyclic processing, determine the final typical PRI value and its left and right boundaries for sorting.

[0013] Furthermore, the method for selecting the initial value of PRI includes:

[0014] S201: Obtain the statistical list; the statistical list includes the numerical element PRI and its corresponding position vector, the total number of pulses vector, and the segment vector;

[0015] S202: Initialize mark as a zero vector of length n; determine the local maximum value and its position, and the local minimum value and its position of the numerical element PRI;

[0016] S203: If the fragment vector corresponding to the local maximum value satisfies the first condition, then the mark record corresponding to the numerical element PRI is 1; otherwise, find the numerical element that is close to the local maximum value between the two local minimum values ​​and record the corresponding mark as 1.

[0017] S204: Find the element that satisfies the second condition from the numerical element PRI, and record the corresponding mark as 1; wherein the second condition includes the total number of pulses vector being greater than the first threshold, or the total number of pulses vector being greater than or equal to the second threshold and the segment vector being greater than or equal to the third threshold;

[0018] S205: Call the PRI uniqueness method continuously until the number of mark elements marked as 1 remains unchanged before and after the two calls. At this time, the positions of the corresponding numerical elements marked as 1 in the vector form the updated vector; remove the updated vector portion from the numerical element vector to update the numerical element vector.

[0019] S206: Traverse the numerical elements PRI to find the element that satisfies the third condition, and mark the mark as 1; the third condition is that the fragment vector is greater than or equal to the fourth threshold, the total number of pulses vector is greater than or equal to the fifth threshold, and / or the feature value is different from the feature value that has been marked with the updated vector within the general tolerance range;

[0020] S207: Call the PRI uniqueness method continuously until the number of mark elements marked as 1 remains unchanged before and after the two calls. At this point, the positions of the numerical elements in the vector marked as 1 form the corresponding vector, which is denoted as the third typical value vector.

[0021] Furthermore, the boundary determination method includes:

[0022] S301: Obtain the statistical list; the statistical list includes the numerical element PRI and its corresponding position vector, the total number of pulses vector, and the segment vector;

[0023] S302: Initialize mark as a zero vector of length n; determine whether the numerical element PRI has been checked for left and right boundaries;

[0024] S303: Check whether the numerical elements marked by the third typical value vector have exceeded the boundary. If not, find the position of the third typical value vector element with the largest corresponding segment vector among all elements of the third typical value vector where the mark element is 0 and store it as the maximum vector.

[0025] S304: Determine the left and right boundaries of the numerical element PRI by determining the difference between the numerical elements.

[0026] Furthermore, the loop processing includes:

[0027] S401: Based on the PRI initial value selection method, a third typical value of PRI is initially selected from the second other numerical elements PRI, and the positions corresponding to the third typical value of PRI constitute the third typical value vector; wherein, the second other numerical elements PRI are composed of the numerical elements PRI recorded in the second other vector;

[0028] If the third typical value vector is not zero, execute S402;

[0029] S402: The boundary determination method is applied to merge the third typical PRI values ​​to obtain the fourth typical PRI value and its corresponding fourth left boundary and fourth right boundary, which are stored in the fourth vector, the fourth left boundary vector and the fourth right boundary vector respectively. The positions of the fourth other PRI values ​​in the numerical elements PRI that are not included by the fourth left boundary and the fourth right boundary are stored in the fourth other vector, and the fourth other vector is sorted from smallest to largest.

[0030] S403: Merge the elements of the second PRI typical value and the fourth PRI typical value and update the second typical value vector to obtain the updated second typical value vector;

[0031] Determine the position of the element of the fourth left boundary vector in the second other vectors, and add the second other typical value to the second left boundary vector to obtain the updated second left boundary vector; determine the position of the element of the fourth right boundary vector in the second other vectors, and add the second other typical value to the second right boundary vector to obtain the updated second right boundary vector;

[0032] S404: Determine the position of the elements of the fourth other typical value in the second other vector, and form the updated second other vector from the data at these positions; if the updated second other vector is not zero, continue to execute S401.

[0033] Furthermore, S106 also includes:

[0034] For each element of the second typical value, add the vector elements of the total number of pulses corresponding to all PRIs represented by the elements of the second left boundary and the second right boundary to obtain the total number of pulses in the PRI typical value interval; compare the pulse count elements of the longest segment corresponding to all PRIs to obtain the pulse count of the longest continuous pulse segment in this PRI typical value interval; when the fifth condition is met, the PRI typical value and interval are deemed valid and used as the final PRI typical value and tolerance; otherwise, the interval is discarded.

[0035] According to a second aspect of the present invention, a histogram-based pulse repetition interval estimation system is provided, comprising:

[0036] Classification module: Used to classify radar reconnaissance signals according to carrier frequency to obtain a dataset for each carrier frequency;

[0037] Histogram calculation module: connected to the classification module, used to draw a histogram based on the data set and generate a statistical list corresponding to the histogram; based on the statistical list, a second typical PRI value is generated by applying the PRI initial value selection method; a second left boundary and a second right boundary are generated corresponding to the second typical PRI value by applying the boundary determination method; the positions of other values ​​to be processed in the statistical list are stored in a second other vector.

[0038] Loop processing module: connected to the histogram calculation module, used to sort the second other vectors from smallest to largest, and perform loop processing until the second other vectors are zero; wherein, the loop processing is to repeatedly execute the PRI initial value selection method and the boundary determination method, and update the typical value of PRI and its left and right boundaries;

[0039] Boundary determination module: Based on the results of the loop processing, determine the final PRI typical value and its left and right boundaries.

[0040] Furthermore, the method for selecting the initial value of PRI includes:

[0041] The method for selecting the initial value of PRI includes:

[0042] S201: Obtain the statistical list; the statistical list includes the numerical element PRI and its corresponding position vector, the total number of pulses vector, and the segment vector;

[0043] S202: Initialize mark as a zero vector of length n; determine the local maximum value and its position, and the local minimum value and its position of the numerical element PRI;

[0044] S203: If the fragment vector corresponding to the local maximum value satisfies the first condition, then the mark record corresponding to the numerical element PRI is 1; otherwise, find the numerical element that is close to the local maximum value between the two local minimum values ​​and record the corresponding mark as 1.

[0045] S204: Find the element that satisfies the second condition from the numerical element PRI, and record the corresponding mark as 1; wherein the second condition includes the total number of pulses vector being greater than the first threshold, or the total number of pulses vector being greater than or equal to the second threshold and the segment vector being greater than or equal to the third threshold;

[0046] S205: Call the PRI uniqueness method continuously until the number of mark elements marked as 1 remains unchanged before and after the two calls. At this time, the positions of the corresponding numerical elements marked as 1 in the vector form the updated vector; remove the updated vector portion from the numerical element vector to update the numerical element vector.

[0047] S206: Traverse the numerical elements PRI to find the element that satisfies the third condition, and mark the mark as 1; the third condition is that the fragment vector is greater than or equal to the fourth threshold, the total number of pulses vector is greater than or equal to the fifth threshold, and / or the feature value is different from the feature value that has been marked with the updated vector within the general tolerance range;

[0048] S207: Call the PRI uniqueness method continuously until the number of mark elements marked as 1 remains unchanged before and after the two calls. At this point, the positions of the numerical elements in the vector marked as 1 form the corresponding vector, which is denoted as the third typical value vector.

[0049] Furthermore, the boundary determination method includes:

[0050] S301: Obtain the statistical list; the statistical list includes the numerical element PRI and its corresponding position vector, the total number of pulses vector, and the segment vector;

[0051] S302: Initialize mark as a zero vector of length n; determine whether the numerical element PRI has been checked for left and right boundaries;

[0052] S303: Check whether the numerical elements marked by the third typical value vector have exceeded the boundary. If not, find the position of the third typical value vector element with the largest corresponding segment vector among all elements of the third typical value vector where the mark element is 0 and store it as the maximum vector.

[0053] S304: Determine the left and right boundaries of the numerical element PRI by determining the difference between the numerical elements.

[0054] Furthermore, the loop processing module is also used for:

[0055] S401: Based on the PRI initial value selection method, a third typical value of PRI is initially selected from the second other numerical elements PRI, and the positions corresponding to the third typical value of PRI constitute the third typical value vector; wherein, the second other numerical elements PRI are composed of the numerical elements PRI recorded in the second other vector;

[0056] If the third typical value vector is not zero, execute S402;

[0057] S402: The boundary determination method is applied to merge the third typical PRI values ​​to obtain the fourth typical PRI value and its corresponding fourth left boundary and fourth right boundary, which are stored in the fourth vector, the fourth left boundary vector and the fourth right boundary vector respectively. The positions of the fourth other PRI values ​​in the numerical elements PRI that are not included by the fourth left boundary and the fourth right boundary are stored in the fourth other vector, and the fourth other vector is sorted from smallest to largest.

[0058] S403: Merge the elements of the second PRI typical value and the fourth PRI typical value and update the second typical value vector to obtain the updated second typical value vector;

[0059] Determine the position of the element of the fourth left boundary vector in the second other vectors, and add the second other typical value to the second left boundary vector to obtain the updated second left boundary vector; determine the position of the element of the fourth right boundary vector in the second other vectors, and add the second other typical value to the second right boundary vector to obtain the updated second right boundary vector;

[0060] S404: Determine the position of the elements of the fourth other typical value in the second other vector, and form the updated second other vector from the data at these positions; if the updated second other vector is not zero, continue to execute S401.

[0061] According to a third aspect of the present invention, a computer-readable storage medium is provided, the computer-readable storage medium storing computer program instructions, by executing the computer program instructions to implement the above-described method and system.

[0062] This invention addresses the shortcomings of existing technologies by proposing a novel histogram-based pulse repetition interval estimation method, system, and storage medium. The method first classifies radar reconnaissance signals according to carrier frequency. For each class, a differential sequence is calculated based on the TOA (Transmission of Alignment) of adjacent pulses. This differential sequence is then processed with a tolerance of 0.1 microseconds. The total number of pulses appearing for each value, the number of pulses contained in the longest consecutive segment of pulse numbers for each value, and the total number of all segments (also known as a "statistical list") are then statistically analyzed. An iterative method is applied to cluster the values ​​in the "statistical list," forming several typical values ​​for repetition intervals (PRI) and their corresponding tolerances. This approach eliminates the need for complex parameter selection, significantly reducing the workload and professional requirements for operators, while simultaneously improving the sorting accuracy.

[0063] The above description is merely an overview of the technical solution of the present invention. In order to better understand the technical means of the present invention and to implement it in accordance with the contents of the specification, the preferred embodiments of the present invention are described in detail below with reference to the accompanying drawings. Attached Figure Description

[0064] The accompanying drawings, which form part of this invention, are used to provide a further understanding of the invention. The illustrative embodiments of the invention and their descriptions are used to explain the invention and do not constitute an undue limitation of the invention. In the drawings:

[0065] Figure 1 A flowchart of the method of the present invention is shown.

[0066] Figure 2 A system block diagram of the present invention is shown.

[0067] Figure 3 A flowchart of another method of the present invention is shown.

[0068] Figure 4 A flowchart of the PRI initial value selection method of the present invention is shown.

[0069] Figure 5 A flowchart of a boundary determination method according to the present invention is shown. Detailed Implementation

[0070] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below in conjunction with specific embodiments and corresponding drawings. Obviously, the described embodiments are only a part of the embodiments of this invention, and not all of them. Based on the embodiments of this invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this invention.

[0071] First, combine Figure 1 The process of this invention is explained. For example... Figure 1 As shown, a pulse repetition interval estimation method based on histograms includes: S101: classifying radar reconnaissance signals according to carrier frequency to obtain a data set for each carrier frequency; S102: drawing a histogram based on the data set and generating a statistical list corresponding to the histogram; S103: generating a second typical PRI value by applying the PRI initial value selection method based on the statistical list; S104: generating a second left boundary and a second right boundary corresponding to the second typical PRI value by applying a boundary determination method, and storing the positions of other values ​​to be processed in the statistical list in a second other vector; S105: sorting the second other vectors from smallest to largest and performing cyclic processing until the second other vector is zero; wherein, the cyclic processing involves repeatedly executing the PRI initial value selection method and the boundary determination method, and updating the typical PRI value and its left and right boundaries; S106: determining the final sorted typical PRI value and its left and right boundaries based on the cyclic processing results.

[0072] Secondly, combine Figure 2 The system of the present invention will be described.

[0073] This invention provides a pulse repetition interval estimation system based on histograms, comprising: a classification module for classifying radar reconnaissance signals according to carrier frequency to obtain a data set for each carrier frequency; a histogram calculation module connected to the classification module for drawing a histogram based on the data set and generating a statistical list corresponding to the histogram; generating a second typical PRI value by applying a PRI initial value selection method based on the statistical list; generating a second left boundary and a second right boundary corresponding to the second typical PRI value by applying a boundary determination method, wherein the positions of other values ​​to be processed in the statistical list are stored in a second other vector; a loop processing module connected to the histogram calculation module for sorting the second other vector from smallest to largest and performing loop processing until the second other vector is zero; wherein the loop processing involves repeatedly executing the PRI initial value selection method and the boundary determination method, and updating the typical PRI value and its left and right boundaries; and a boundary determination module connected to the loop processing module for determining the final sorted typical PRI value and its left and right boundaries based on the loop processing results.

[0074] The following is combined Figure 3 Provide a detailed description.

[0075] A pulse repetition interval estimation method based on histograms, the pulse repetition interval estimation method comprising the following steps:

[0076] Step S101: Classify radar reconnaissance signals according to carrier frequency to obtain a data set for each carrier frequency.

[0077] Step S102: Draw a histogram based on the data set and generate a statistical list corresponding to the histogram.

[0078] Specifically, for each category, a histogram is calculated with a tolerance of 0.1 microseconds. This involves calculating the frequency of each value (i.e., the total number of pulses) based on the Time of Arrival (TOA) by subtracting adjacent values. Simultaneously, the number of pulses contained in the longest consecutive segment of pulse numbers for each value is analyzed. The histogram values ​​are sorted in ascending order to obtain the vector `pri`. The total number of pulses for each value forms the vector `SeqTotalLen`, and the number of pulses contained in the longest segment forms the vector `MaxSeqLen`.

[0079] Step S103: Based on the statistical list, generate a second typical PRI value using the PRI initial value selection method.

[0080] Specifically, the "PRI initial value selection method" autoSelePri2 is applied to obtain the vector index6, which is composed of the positions of the typical values ​​initially selected from the pri vector;

[0081] Step S104: Apply the boundary determination method to generate the second left boundary and the second right boundary corresponding to the second typical PRI value. The positions of other PRI values ​​to be processed in the statistical list are stored in the second other vector.

[0082] Specifically, the "boundary determination method" findBound is applied to further merge the typical values ​​initially selected by index6 to obtain new typical values ​​and their corresponding left and right boundaries, which are stored in the vectors index8, index7_L, and index7_R, respectively. The positions of other values ​​to be processed in pri that are not included by these boundaries are stored in the vector index9 (the second other vector).

[0083] Step S105: Sort the second other vectors from smallest to largest, and process them in a loop until the second other vectors are zero.

[0084] Specifically, sort the integer vector index9 in ascending order. Repeat steps S1051 to S1054 until index9 is empty.

[0085] Step S1051: Form a vector from the pri elements recorded in index9, denoted as pri3; form a vector from the corresponding position vector, denoted as index3; form a vector from the corresponding MaxSeqLen elements, denoted as MaxSeqLen3; form a vector from the corresponding SeqTotalLen elements, denoted as SeqTotalLen3; apply the "PRI initial value selection method" autoSelePri2 to obtain the position vector index99 corresponding to the typical value initially selected from pri3; if index99 is empty, proceed to step nine; otherwise, proceed to step six.

[0086] Step S1052: Apply the "determine boundary method" findBound to further merge the typical values ​​initially selected by index99, and obtain new selected typical values ​​and their corresponding left and right boundaries, which are stored in the vectors index11, index10_L, and index10_R respectively. The positions of other values ​​to be processed in pri that are not included by these boundaries are stored in the vector index13 (the updated second other vector). Then, sort the integer vector index13 from smallest to largest.

[0087] Step S1053: Add each element of index11 to index8 and save it as a new index8 vector. Find the position of the element of index10_L in index9, add the data of these positions of index9 to index7_L, and save it as a new index7_L vector. Find the position of the element of index10_R in index9, add the data of these positions of index9 to index7_R, and save it as a new index7_R vector.

[0088] Step S1054: Find the position of the element at index13 in index9, and form a new vector at index9 using the data at these positions. If index9 is not empty, continue to step five; otherwise, proceed to step nine.

[0089] Step S106, S106 further includes:

[0090] For each element of the second PRI typical value, add the corresponding pulse count vector elements of all PRI corresponding to the elements of the second left boundary and the second right boundary to obtain the total number of pulses in the PRI typical value interval; compare the pulse count elements of the longest segment corresponding to all PRI to obtain the pulse count of the longest continuous pulse segment in this PRI typical value interval; when the fifth condition is met, the PRI typical value and interval are deemed valid and used as the last sorted PRI typical value and tolerance (left boundary, right boundary); otherwise, the interval is discarded.

[0091] Specifically, for each element of index8, the SeqTotalLen elements corresponding to all pri values ​​represented by the elements of index7_L and index7_R are summed to obtain the total number of pulses within this typical PRI value interval, denoted as sumPul. The MaxSeqLen elements corresponding to all pri values ​​are compared to obtain the number of pulses in the longest continuous pulse segment within this typical PRI value interval, denoted as maxLen. When a certain condition is met, the typical PRI value and interval are considered valid and are used as the final sorted typical PRI value and tolerance (left boundary, right boundary); otherwise, the interval is discarded.

[0092] Furthermore, the method for selecting the initial value of PRI includes:

[0093] S201: Obtain the statistical list; the statistical list includes the numerical element PRI and its corresponding position vector, the total number of pulses vector, and the segment vector;

[0094] S202: Initialize mark as a zero vector of length n; determine the local maximum value and its position, and the local minimum value and its position of the numerical element PRI;

[0095] S203: If the fragment vector corresponding to the local maximum value satisfies the first condition, then the mark record corresponding to the numerical element PRI is 1; otherwise, find the numerical element that is close to the local maximum value between the two local minimum values ​​and record the corresponding mark as 1.

[0096] S204: Find the element that satisfies the second condition from the numerical element PRI, and record the corresponding mark as 1; wherein the second condition includes the total number of pulses vector being greater than the first threshold, or the total number of pulses vector being greater than or equal to the second threshold and the segment vector being greater than or equal to the third threshold;

[0097] S205: Call the PRI uniqueness method continuously until the number of mark elements marked as 1 remains unchanged before and after the two calls. At this time, the positions of the corresponding numerical elements marked as 1 in the vector form the updated vector; remove the updated vector portion from the numerical element vector to update the numerical element vector.

[0098] S206: Traverse the numerical elements PRI to find the element that satisfies the third condition, and mark the mark as 1; the third condition is that the fragment vector is greater than or equal to the fourth threshold, the total number of pulses vector is greater than or equal to the fifth threshold, and / or the feature value is different from the feature value that has been marked with the updated vector within the general tolerance range;

[0099] S207: Call the PRI uniqueness method continuously until the number of mark elements marked as 1 remains unchanged before and after the two calls. At this point, the positions of the numerical elements in the vector marked as 1 form the corresponding vector, which is denoted as the third typical value vector.

[0100] The following is combined Figure 4 A more detailed description of the method for selecting the initial value of PRI is provided.

[0101] Step S201: Obtain the vector pri3, the position vector index3, the counting vector SeqTotalLen3, and the length vector MaxSeqLen3, which are composed of the elements corresponding to the repeated interval feature values ​​to be evaluated in ascending order.

[0102] Step S202: Let the length of pri3 be n. Initialize mark as a zero vector of length n. Find the positions of the local maxima and local minima in pri3.

[0103] Step S203: For the pri3 element of the local maximum value, if the corresponding MaxSeqLen3 value satisfies a certain condition, then the mark element corresponding to the position of the element is marked as 1. Otherwise, between the two local minimum positions before and after it, search for other pri3 elements that are close to the value of the local maximum value (within the general tolerance range) to see if there is an element that satisfies a certain condition. If there is, then the mark element corresponding to the above local maximum value is also marked as 1.

[0104] Step S204: Find the element in the pri3 vector that satisfies "SeqTotalLen3 is greater than a certain threshold" or "MaxSeqLen3 is greater than or equal to a certain threshold and SeqTotalLen3 is greater than or equal to a certain threshold", and mark the corresponding mark element as 1.

[0105] Step S205: Call the PRI uniqueness method unique_pri continuously until the number of mark elements marked as 1 remains unchanged before and after the two calls. At this time, the positions of those pri3 vectors marked as 1 form a vector, denoted as index4; remove the index4 part from index3 and update index3.

[0106] Step S206: Traverse the pri3 elements corresponding to each index3 element, and find the elements that simultaneously satisfy "MaxSeqLen3 is greater than or equal to a certain threshold", "SeqTotalLen3 is greater than or equal to a certain threshold", and "are different from the feature values ​​already marked by index4 within the general tolerance range". The corresponding mark is set to 1.

[0107] Step S207: Call the PRI uniqueness method unique_pri continuously until the number of mark elements marked as 1 remains unchanged before and after the two calls. At this time, the positions of those pri3 vectors marked as 1 form a vector, denoted as index99 (the third typical value vector).

[0108] Furthermore, the boundary determination method includes:

[0109] S301: Obtain the statistical list; the statistical list includes the numerical element PRI and its corresponding position vector, the total number of pulses vector, and the segment vector;

[0110] S302: Initialize mark as a zero vector of length n; determine whether the numerical element PRI has been checked for left and right boundaries;

[0111] S303: Check whether the numerical elements marked by the third typical value vector have exceeded the boundary. If not, find the position of the third typical value vector element with the largest corresponding segment vector among all elements of the third typical value vector where the mark element is 0 and store it as the maximum vector.

[0112] S304: Determine the left and right boundaries of the numerical element PRI by determining the difference between the numerical elements.

[0113] The following is combined Figure 5 The boundary determination method is described in more detail below.

[0114] Step S301: Obtain the vector pri3 composed of the repeating interval feature values ​​to be evaluated, sorted from smallest to largest; the vector index99 composed of the positions of the potential PRI typical values ​​in the pri3 vector (i.e., the output of the "PRI initial value selection method" autoSelePri2); the counting vector SeqTotalLen3 composed of the corresponding elements in the initial counting vector SeqTotalLen; and the length vector MaxSeqLen3 composed of the corresponding elements in the initial length vector MaxSeqLen.

[0115] Step S302: Let the length of pri3 be n. Initialize mark as a zero vector of length n. Mark the corresponding elements of pri3 to indicate whether the left and right boundaries have been checked. A value of 1 indicates that they have been checked, and a value of 0 indicates that they have not been checked.

[0116] Step S303: Check if all the corresponding mark elements of the corresponding element of pri3 marked index99 are 1. If yes, the method ends. If not, find the position of the index99 element with the largest MaxSeqLen among all the index99 elements with mark element 0, and denot it as maxindex.

[0117] Step S304: Determine the left and right boundaries of the numerical element PRI by determining the differences between the numerical elements. This step specifically includes:

[0118] Repeat steps S3041 to S30410.

[0119] Step S3041: When maxindex is 0, corresponding to the first element in index99, start from pri3[index99[maxindex]] and search leftwards for the farthest element that satisfies the condition that the difference between pri3[index99[maxindex]] and the element is less than a certain condition. Set its position as the left boundary of the typical value of pri3 and record it in index10_L; proceed to step S3049. When maxindex is greater than 0, analyze whether the difference between the elements corresponding to index99[maxindex] and index99[maxindex-1] in pri3 is greater than the above condition. If so, proceed to step S3042; otherwise, proceed to step S3044.

[0120] Step S3042: When mark[index99[maxindex-1]] is 1, proceed to step S3043; when mark[index99[maxindex-1]] is 0, examine each element between pri3[index99[maxindex-1]+1] and pri3[index99[maxindex]-1], find those elements whose difference with pri3[index99[maxindex]] is less than or equal to the condition in step S3041, and whose corresponding MaxSeqLen value is less than a certain threshold or whose corresponding SeqTotalLen value is less than a certain threshold, determine the position of these elements that is farthest from index99[maxindex], set its position as the left boundary of the typical value of pri3 marked by index99[maxindex], and record it in index10_L. Proceed to step S3049. If no element meets the condition, proceed to step A1.1;

[0121] Step A1.1: Find the right boundary of the previous typical value pri3[index99[maxindex-1]]: Examine each element between pri3[index99[maxindex-1]+1] and pri3[index99[maxindex]-1]. Find those elements whose difference from pri3[index99[maxindex-1]] is less than or equal to the condition in step S3041, and whose corresponding MaxSeqLen value is less than a certain threshold or whose corresponding SeqTotalLen value is less than a certain threshold. Determine the position of these elements that is farthest from index99[maxindex-1]. If the difference between this element and pri3[index99[maxindex]] is also less than or equal to the condition in step S3041, then its position is set as the left boundary of the typical value of pri3 marked by index99[maxindex], and recorded in index10_L. Proceed to step 12. If the above elements do not exist, proceed to step A1.2;

[0122] Step A1.2: Starting from pri3[index99[maxindex]], search to the left for the farthest element that satisfies the condition in step S3041 where the difference between pri3[index99[maxindex]] and the element is less than the value in step S3041. Set its position to the left boundary of the typical value of pri3 marked by index99[maxindex], and record it in index10_L. Proceed to step S3049.

[0123] Step S3043: Examine whether the difference between pri3[index99[maxindex]] and pri3[index10_R[maxindex-1]+1] is less than or equal to the condition in step S3041. If yes, the left boundary of the typical value of pri3[index99[maxindex]] is index10_R[maxindex-1]+1, and it is recorded in index10_L; proceed to step S3049. If no, starting from pri3[index99[maxindex]], search to the left for the farthest element that satisfies the condition in step S3041 that the difference with pri3[index99[maxindex]] is less than the condition in step S3041, and set its position as the left boundary of the typical value of pri3, and record it in index10_L; proceed to step S3049.

[0124] Step S3044: When maxindex is 1, starting from pri3[index99[maxindex]], search to the left for the farthest element that satisfies the condition in step S3041 where the difference between pri3[index99[maxindex]] and maxindex is less than the difference in index10_L. Set its position to the left boundary of that typical value of pri3 and record it in index10_L; proceed to step S3049. When maxindex is greater than 1, execute step S3045.

[0125] Step S3045: Analyze whether the difference between the corresponding elements of index99[maxindex] and index99[maxindex-2] in pri3 is greater than the condition in step S3041. If so, proceed to step S3046; otherwise, proceed to step S3049.

[0126] Step S3046: When mark[index99[maxindex-1]] is 1, proceed to step S3047; when mark[index99[maxindex-1]] is 0 and mark[index6[maxindex-2]] is 1, proceed to step S3048; otherwise, examine each element between pri3[index99[maxindex-2]+1] and pri3[index99[maxindex]-1], find those elements whose difference with pri3[index99[maxindex]] is less than or equal to the condition in step S3041, and whose corresponding MaxSeqLen value is less than a certain threshold or whose corresponding SeqTotalLen value is less than a certain threshold, determine the position of these elements that is farthest from index99[maxindex], set its position as the left boundary of the typical value of pri3 marked by index99[maxindex], and record it in index10_L. Proceed to step S3049. If the element that meets the condition does not exist, proceed to step B1.1;

[0127] Step B1.1: Find the right boundary of the typical value pri3[index99[maxindex-2]]: Examine each element between pri3[index99[maxindex-2]+1] and pri3[index99[maxindex]-1]. Find those elements whose difference from pri3[index99[maxindex-2]] is less than or equal to the condition in step S3041, and whose corresponding MaxSeqLen value is less than a certain threshold or whose corresponding SeqTotalLen value is less than a certain threshold. Determine the position of these elements that is farthest from index99[maxindex-2]. If the difference between this element and pri3[index99[maxindex]] is also less than or equal to the condition in step S3041, then its position is set as the left boundary of the typical value of pri3 marked by index99[maxindex], and recorded in index10_L. Proceed to step S3049. If the above element does not exist, proceed to step B1.2;

[0128] Step B1.2: Starting from pri3[index99[maxindex]], search to the left for the farthest element that satisfies the condition in step S3041 where the difference between pri3[index99[maxindex]] and the element is less than the value in step S3041. Set its position to the left boundary of the typical value of pri3 marked by index99[maxindex], and record it in index10_L. Proceed to step S3049.

[0129] Step S3047: Examine whether the difference between pri3[index99[maxindex]] and pri3[index10_R[maxindex-1]+1] is less than or equal to the condition in step S3041. If yes, the left boundary of the typical value of pri3[index99[maxindex]] is index10_R[maxindex-1]+1, and it is recorded in index10_L; proceed to step S3049. If no, starting from pri3[index99[maxindex]], search to the left for the farthest element that satisfies the condition in step S3041 that the difference with pri3[index99[maxindex]] is less than the condition in step S3041, and set its position as the left boundary of that typical value of pri3, and record it in index10_L. Proceed to step S3049.

[0130] Step S3048: Examine whether the difference between pri3[index99[maxindex]] and pri3[index10_R[maxindex-2]+1] is less than or equal to the condition in step S3041. If yes, the left boundary of the typical value of pri3[index99[maxindex]] is index10_R[maxindex-2]+1, and it is recorded in index10_L; proceed to step S3049. If no, starting from pri3[index99[maxindex]], search to the left for the farthest element that satisfies the condition in step S3041 that the difference with pri3[index99[maxindex]] is less than the condition in step S3041, and set its position as the left boundary of that typical value of pri3, and record it in index10_L. Proceed to step S3049.

[0131] Step S3049: Following the steps S3041 to S3048, execute the code to determine the right boundary of the typical value of pri3[index99[maxindex]] and record it in index10_R. Proceed to step S30410.

[0132] Step S30410: If index99[maxindex-1] is located within the closed interval between index10_L and index10_R, then the left and right boundaries of pri3[index99[maxindex-1]] are also the same value, and the same applies to index99[maxindex+1]. All elements in the mark array located between index10_L and index10_R are assigned the value 1. Then, proceed to step S303 and repeat the loop until the left and right boundaries of the last element of index99 are also determined.

[0133] Step S30411: Arrange the positions of the elements in the pri3 vector that are not contained within the boundary specified by any pair of index10_L and index10_R elements in order, and denote them as the remaining vector index13.

[0134] Furthermore, the loop processing includes:

[0135] S401: Based on the PRI initial value selection method, a third typical value of PRI is initially selected from the second other numerical elements PRI, and the positions corresponding to the third typical value of PRI constitute the third typical value vector; wherein, the second other numerical elements PRI are composed of the numerical elements PRI recorded in the second other vector;

[0136] If the third typical value vector is not zero, execute S402;

[0137] S402: The boundary determination method is applied to merge the third typical PRI values ​​to obtain the fourth typical PRI value and its corresponding fourth left boundary and fourth right boundary, which are stored in the fourth vector, the fourth left boundary vector and the fourth right boundary vector respectively. The positions of the fourth other PRI values ​​in the numerical elements PRI that are not included by the fourth left boundary and the fourth right boundary are stored in the fourth other vector, and the fourth other vector is sorted from smallest to largest.

[0138] S403: Merge the elements of the second PRI typical value and the fourth PRI typical value and update the second typical value vector to obtain the updated second typical value vector;

[0139] Determine the position of the element of the fourth left boundary vector in the second other vectors, and add the second other typical value to the second left boundary vector to obtain the updated second left boundary vector; determine the position of the element of the fourth right boundary vector in the second other vectors, and add the second other typical value to the second right boundary vector to obtain the updated second right boundary vector;

[0140] S404: Determine the position of the elements of the fourth other typical value in the second other vector, and form the updated second other vector from the data at these positions; if the updated second other vector is not zero, continue to execute S401.

[0141] The beneficial effects of this invention are:

[0142] 1. This invention first classifies radar reconnaissance signals according to carrier frequency, forming a data set for each carrier frequency. Then, it uses histogram statistics of adjacent pulse TOA differential sequences as sample points with a tolerance of 0.1 microseconds. The total number of pulses and the length of the longest pulse segment are used as parameters to obtain the typical PRI value under each carrier frequency data set, while obtaining their respective different tolerances. This avoids the shortcomings of using a fixed tolerance for PRI sorting and improves the sorting accuracy.

[0143] 2. The classification process of this invention can be viewed as clustering: the vast majority of sample points are clustered into some typical classes, each class corresponding to a typical PRI value and its range. However, for clustering algorithms, the "total number of pulses" and "longest pulse segment length" of sample points are unrelated. To obtain reasonable clustering results, it is necessary to pre-set the "maximum allowable uncovered sample rate," which is not suitable for radar signal sorting in complex situations. This invention focuses on the specific meaning of "total number of pulses" and "longest pulse segment length," and through the iterative use of the "PRI initial value selection method" and the "boundary determination method," the classification process is continuously expanded until a more reasonable typical PRI value and range are determined. The subsequent classification result is always better than the previous classification result. It is particularly suitable for radar signal classification with a large sample size but significant disturbances and measurement errors.

[0144] 3. Compared with radar signal sorting algorithms such as CDIF, SDIF, and PRI transform, this invention provides an alternative sorting approach. Based on the differential sequence of all sample points at 0.1 microsecond intervals, it rapidly and stably obtains the typical PRI value and variable tolerance through continuous iteration.

[0145] 4. This invention does not require complex parameter selection, which greatly reduces the workload and professional requirements of operators.

[0146] It should be noted that, unless otherwise specified, the embodiments and features described in this invention can be combined with each other. The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention in any way. Any simple modifications, equivalent changes, and alterations made to the above embodiments based on the technical essence of the present invention shall still fall within the scope of the present invention.

[0147] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the embodiments of the present invention and are not intended to limit them. Although the embodiments of the present invention have been described in detail with reference to the above preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions to the technical solutions of the embodiments of the present invention should not depart from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A pulse repetition interval estimation method based on histograms, characterized in that, include: S101: Classify radar reconnaissance signals according to carrier frequency to obtain a dataset for each carrier frequency; S102: Draw a histogram based on the data set and generate a statistical list corresponding to the histogram; S103: Based on the statistical list, the second typical PRI value is generated by applying the PRI initial value selection method; S104: Apply the boundary determination method to generate the second left boundary and the second right boundary corresponding to the second typical PRI value, and store the positions of other PRI values ​​to be processed in the statistical list in the second other vector; S105: Sort the second other vector in ascending order, and process it in a loop until the second other vector is zero; The loop processing involves repeatedly executing the PRI initial value selection method and the boundary determination method, and updating the typical PRI value and its left and right boundaries. S106: Based on the results of the cyclic processing, determine the final typical PRI value and its left and right boundaries for sorting; The method for selecting the initial value of PRI includes: S201: Obtain the statistical list; the statistical list includes the numerical element PRI and its corresponding position vector, the total number of pulses vector, and the segment vector; S202: Initialize mark as a zero vector of length n; determine the local maximum value and its position, and the local minimum value and its position of the numerical element PRI; S203: If the fragment vector corresponding to the local maximum value satisfies the first condition, then the mark record corresponding to the numerical element PRI is 1; otherwise, search for a numerical element close to the local maximum value between the two local minimum values ​​and record the corresponding mark as 1; wherein the first condition includes the fragment vector value corresponding to the local maximum value meeting the preset filtering requirements. S204: Find the element that satisfies the second condition from the numerical element PRI, and record the corresponding mark as 1; wherein the second condition includes the total number of pulses vector being greater than the first threshold, or the total number of pulses vector being greater than or equal to the second threshold and the segment vector being greater than or equal to the third threshold; S205: Call the PRI uniqueness method continuously until the number of mark elements marked as 1 remains unchanged before and after the two calls. At this time, the positions of the corresponding numerical elements marked as 1 in the vector form the updated vector; remove the updated vector portion from the numerical element vector to update the numerical element vector. S206: Traverse the numerical elements PRI to find the element that satisfies the third condition, and mark the mark as 1; the third condition is that the fragment vector is greater than or equal to the fourth threshold, the total number of pulses vector is greater than or equal to the fifth threshold, and / or the feature value is different from the feature value that has been marked with the updated vector within the general tolerance range; S207: Call the PRI uniqueness method continuously until the number of mark elements marked as 1 remains unchanged before and after the two calls. At this point, the positions of the numerical elements in the vector marked as 1 form the corresponding vector, which is denoted as the third typical value vector.

2. The method as described in claim 1, characterized in that, The boundary determination method includes: S301: Obtain the statistical list; the statistical list includes the numerical element PRI and its corresponding position vector, the total number of pulses vector, and the segment vector; S302: Initialize mark as a zero vector of length n; determine whether the numerical element PRI has been checked for left and right boundaries; S303: Check whether the numerical elements marked by the third typical value vector have exceeded the boundary. If not, find the position of the third typical value vector element with the largest corresponding segment vector among all elements of the third typical value vector where the mark element is 0 and store it as the maximum vector. S304: Determine the left and right boundaries of the numerical element PRI by determining the difference between the numerical elements.

3. The method as described in claim 2, characterized in that, The loop process includes: S401: Based on the PRI initial value selection method, a third typical value of PRI is initially selected from the second other numerical elements PRI composed of the numerical elements PRI recorded in the second other vector, and the position corresponding to the third typical value of PRI constitutes the third typical value vector. If the third typical value vector is not zero, execute S402; S402: The boundary determination method is applied to merge the third typical PRI values ​​to obtain the fourth typical PRI value and its corresponding fourth left boundary and fourth right boundary, which are stored in the fourth vector, the fourth left boundary vector and the fourth right boundary vector respectively. The positions of the fourth other PRI values ​​in the numerical elements PRI not included by the fourth left boundary and the fourth right boundary are stored in the fourth other vector, and the fourth other vector is sorted from smallest to largest. S403: Merge the elements of the second PRI typical value and the fourth PRI typical value and update the second typical value vector to obtain the updated second typical value vector; Determine the position of the element of the fourth left boundary vector in the second other vectors, and add the second other typical value to the second left boundary vector to obtain the updated second left boundary vector; determine the position of the element of the fourth right boundary vector in the second other vectors, and add the second other typical value to the second right boundary vector to obtain the updated second right boundary vector; S404: Determine the position of the elements of the fourth other typical value in the second other vector, and form the updated second other vector from the data at these positions; if the updated second other vector is not zero, continue to execute S401.

4. The method as described in claim 3, characterized in that, S106 further includes: For each element of the second typical value, add the vector elements of the total number of pulses corresponding to all PRIs represented by the elements of the second left boundary and the second right boundary to obtain the total number of pulses in the PRI typical value interval; compare the pulse count elements of the longest segment corresponding to all PRIs to obtain the pulse count of the longest continuous pulse segment in this PRI typical value interval; when the fifth condition is met, the PRI typical value and interval are deemed valid and used as the final PRI typical value and tolerance; otherwise, the interval is discarded.

5. A pulse repetition interval estimation system based on histograms, characterized in that, include: Classification module: Used to classify radar reconnaissance signals according to carrier frequency to obtain a dataset for each carrier frequency; Histogram calculation module: connected to the classification module, used to draw a histogram based on the data set and generate a statistical list corresponding to the histogram; based on the statistical list, a second typical PRI value is generated by applying the PRI initial value selection method; a second left boundary and a second right boundary are generated corresponding to the second typical PRI value by applying the boundary determination method; the positions of other values ​​to be processed in the statistical list are stored in a second other vector. Loop processing module: connected to the histogram calculation module, used to sort the second other vectors from smallest to largest, and perform loop processing until the second other vectors are zero; wherein, the loop processing is to repeatedly execute the PRI initial value selection method and the boundary determination method, and update the typical value of PRI and its left and right boundaries; Boundary determination module: Based on the results of the loop processing, determine the final typical PRI value and its left and right boundaries; The method for selecting the initial value of PRI includes: The method for selecting the initial value of PRI includes: S201: Obtain the statistical list; the statistical list includes the numerical element PRI and its corresponding position vector, the total number of pulses vector, and the segment vector; S202: Initialize mark as a zero vector of length n; determine the local maximum value and its position, and the local minimum value and its position of the numerical element PRI; S203: If the fragment vector corresponding to the local maximum value satisfies the first condition, then the mark record corresponding to the numerical element PRI is 1; otherwise, search for a numerical element close to the local maximum value between the two local minimum values ​​and record the corresponding mark as 1; wherein the first condition includes the fragment vector value corresponding to the local maximum value meeting the preset filtering requirements. S204: Find the element that satisfies the second condition from the numerical element PRI, and record the corresponding mark as 1; wherein the second condition includes the total number of pulses vector being greater than the first threshold, or the total number of pulses vector being greater than or equal to the second threshold and the segment vector being greater than or equal to the third threshold; S205: Call the PRI uniqueness method continuously until the number of mark elements marked as 1 remains unchanged before and after the two calls. At this time, the positions of the corresponding numerical elements marked as 1 in the vector form the updated vector; remove the updated vector portion from the numerical element vector to update the numerical element vector. S206: Traverse the numerical elements PRI to find the element that satisfies the third condition, and mark the mark as 1; the third condition is that the fragment vector is greater than or equal to the fourth threshold, the total number of pulses vector is greater than or equal to the fifth threshold, and / or the feature value is different from the feature value that has been marked with the updated vector within the general tolerance range; S207: Call the PRI uniqueness method continuously until the number of mark elements marked as 1 remains unchanged before and after the two calls. At this point, the positions of the numerical elements in the vector marked as 1 form the corresponding vector, which is denoted as the third typical value vector.

6. The system as described in claim 5, characterized in that, The boundary determination method includes: S301: Obtain the statistical list; the statistical list includes the numerical element PRI and its corresponding position vector, the total number of pulses vector, and the segment vector; S302: Initialize mark as a zero vector of length n; determine whether the numerical element PRI has been checked for left and right boundaries; S303: Check whether the numerical elements marked by the third typical value vector have exceeded the boundary. If not, find the position of the third typical value vector element with the largest corresponding segment vector among all elements of the third typical value vector where the mark element is 0 and store it as the maximum vector. S304: Determine the left and right boundaries of the numerical element PRI by determining the difference between the numerical elements.

7. The system as described in claim 6, characterized in that, The loop processing module is also used for: S401: Based on the PRI initial value selection method, a third typical value of PRI is initially selected from the second other numerical elements PRI composed of the numerical elements PRI recorded in the second other vector, and the position corresponding to the third typical value of PRI constitutes the third typical value vector. If the third typical value vector is not zero, execute S402; S402: The boundary determination method is applied to merge the third typical PRI values ​​to obtain the fourth typical PRI value and its corresponding fourth left boundary and fourth right boundary, which are stored in the fourth vector, the fourth left boundary vector and the fourth right boundary vector respectively. The positions of the fourth other PRI values ​​in the numerical elements PRI not included by the fourth left boundary and the fourth right boundary are stored in the fourth other vector, and the fourth other vector is sorted from smallest to largest. S403: Merge the elements of the second PRI typical value and the fourth PRI typical value and update the second typical value vector to obtain the updated second typical value vector; Determine the position of the element of the fourth left boundary vector in the second other vectors, and add the second other typical value to the second left boundary vector to obtain the updated second left boundary vector; determine the position of the element of the fourth right boundary vector in the second other vectors, and add the second other typical value to the second right boundary vector to obtain the updated second right boundary vector; S404: Determine the position of the elements of the fourth other typical value in the second other vector, and form the updated second other vector from the data at these positions; if the updated second other vector is not zero, continue to execute S401.

8. A computer-readable storage medium storing computer program instructions, wherein executing the computer program instructions implements the method of any one of claims 1-4.

Images