A method for combining radiation source pulses based on inter-frame waveform similarity calculation

Abstract

The present application belongs to the technical field of radar data processing and signal sorting, and relates to a radiation source pulse combination method based on inter-frame waveform similarity calculation. The method realizes effective combination of complex radar radiation source pulse groups by clustering intercepted pulse groups, extracting timing rules and calculating waveform similarity, and solves the problem of difficult batch sorting of complex radar signals. The implementation steps are as follows: all pulse groups intercepted in the current sorting beat are traversed, and pulse group classes are generated based on pulse group parameters; for the newly generated pulse group classes, the pulse groups contained therein are sorted according to arrival time and timing rules are extracted; the timing rules in the pulse group classes are counted and updated; for the pulse group classes generated in different sorting beats, the waveform similarity is calculated by comparing the timing rules contained in the pulse group classes, and the pulse group class combination is completed. The present application is conducive to improving the batch accuracy of complex radar radiation sources, and can be used in radar detection systems.

Description

Technical Field

[0001] This invention belongs to the field of radar data processing and signal sorting technology. Background Technology

[0002] For complex radar systems, the signals intercepted by the receiver are typically in the form of pulse groups, each containing several to dozens of consecutive pulses. These pulse groups usually have similar frequencies, pulse widths, and amplitudes, and the arrival time differences between adjacent pulses satisfy a specific modulation pattern. In engineering implementation, the intercepted radiation source pulse data is typically sorted based on a preset cycle, and an EDW (Earning Rendering) is generated for each cycle. The pulse groups emitted by complex radar systems at different wave positions may employ different repetition period values ​​or even different repetition period types, resulting in EDWs generated at different cycles potentially having different repetition period types and values.

[0003] Effectively merging EDWs with different repetition period values ​​and even different types to reduce the batching of radiation sources is a technical challenge. Analysis of a large amount of actual intercepted data reveals that the repetition period values ​​of complex radar systems, typically within a single frame, exhibit certain regularities within a certain time range. Therefore, there is an urgent need to propose a radiation source pulse merging method based on inter-frame waveform similarity calculation to improve the batching accuracy of this type of radiation source. Summary of the Invention

[0004] The purpose of this invention is to provide a method for merging radiation source pulse groups based on inter-frame waveform similarity calculation. By clustering the intercepted pulse groups, extracting timing rules, and calculating waveform similarity, the method can effectively merge complex radiation source pulse groups and improve the accuracy of merging radiation sources of complex radar systems.

[0005] To achieve the above technical objectives, the present invention adopts the following technical solution:

[0006] A method for combining radiation source pulses based on inter-frame waveform similarity calculation includes the following steps:

[0007] Step 1: Traverse the pulse groups intercepted within the current sorting cycle and generate pulse group classes based on pulse group parameters: For any two pulse groups with a fixed repetition period type, determine whether the two pulse groups overlap in the time domain. If there is no overlap, further determine whether the repetition period values ​​of the two are the same. If they are the same, determine that the two belong to the same class. If they are different, calculate the arrival time difference between the two and set a time threshold T1 to determine whether the two belong to the same class.

[0008] Step 2: For the newly generated pulse group class C k Sort all pulse groups contained therein according to their arrival time and extract timing rules: Traverse C kFor pulse groups, for adjacent pulse groups, calculate the arrival time difference between them; if the arrival time difference is within a preset time threshold T2, then generate a timing rule r = p pre →p suf , where p pre p is the repetition period value of the previous pulse group. suf This is the repetition period value of the next pulse group;

[0009] Step 3: Statistically analyze and update pulse group class C k Timing rules in: For a newly generated timing rule r, match r with pulse group class C k The set of time-series rules R included k All time series rules in the set are compared pairwise to determine if they are identical. If identical time series rules are found, the count of the corresponding time series rule is incremented by one; otherwise, r is added to the set R. k middle;

[0010] Step 4: Generate two pulse group classes C for different sorting cycles. m and C n The time series rules it contains are R m and R n Calculate C m and C n waveform similarity Where N eq For pulse group class C m and C n The number of identical timing rules contained, N m For R m The number of time-series rules included, N n For R n The number of timing rules included; if the waveform similarity is greater than a set threshold, then C is determined. m and C n Those belonging to the same radiation source are merged, and C is... m and C n All pulse groups included are added to the merged pulse group class C. F middle.

[0011] The beneficial effects of this invention are as follows: In view of the problem that it is difficult to merge signals of complex radar radiation sources when there is no prior information, a radiation source pulse group merging method based on inter-frame waveform similarity calculation is proposed. By clustering all captured pulse groups, extracting timing rules and calculating waveform similarity, the pulse groups are effectively merged, thereby improving the batching accuracy of complex radar radiation sources. Attached Figure Description

[0012] Figure 1This is a flowchart of a radiation source pulse combination method based on inter-frame waveform similarity calculation according to the present invention. Detailed Implementation

[0013] Figure 1 This is a flowchart of a radiation source pulse combination method based on inter-frame waveform similarity calculation according to the present invention. Referring to the accompanying drawings, the specific steps of the present invention are as follows:

[0014] Step 1: Traverse all pulse groups captured within the current sorting cycle, and generate pulse group classes based on pulse group parameters:

[0015] (1.1) For any two pulse groups with a fixed repetition period, determine whether the two pulse groups overlap in the time domain. If the arrival time of the pulses contained in the two pulse groups satisfies the formula (1), then it is determined that the two do not overlap.

[0016] TOA f -TOA l >T d | (1)

[0017] Among them, TOA f TOA represents the arrival time of the first pulse in the next pulse group. l T represents the arrival time of the last pulse in the previous pulse group. d This indicates the preset time difference threshold;

[0018] (1.2) If the two pulse groups do not overlap in the time domain, further determine whether the repetition period of the two pulse groups is the same;

[0019] (1.3) If the repetition period of two pulse groups is the same, they are determined to belong to the same category. If they are different, the arrival time difference pgdTOA is calculated according to formula (2). If the arrival time difference is less than or equal to the set time threshold T1, they are determined to belong to the same category.

[0020] pgdTOA=TOA i -TOA j (2)

[0021] Among them, TOA i TOA represents the arrival time of the first pulse in the next pulse group. j This indicates the arrival time of the first pulse in the previous pulse group;

[0022] Step 2: For the newly generated pulse group class C k Sort all pulse groups contained therein according to their arrival time and extract timing rules: Traverse C kFor pulse groups, for adjacent pulse groups, calculate the arrival time difference between them; if the arrival time difference is within a preset time threshold T2, then generate a timing rule r = p pre →p suf , where p pre p is the repetition period value of the previous pulse group. suf This is the repetition period value of the next pulse group;

[0023] Step 3: Statistically analyze and update pulse group class C k Timing rules in:

[0024] (3.1) For the newly generated timing rule r, match r with the pulse group class C k The set of time-series rules R included k Compare all the time series rules in the data, for the time series rule r = p pre →p suf and R k any one of the timing rules If the conditions described in formula (3) are met, then the timing rules r and r are considered to be true. p same;

[0025]

[0026] Where t1 and t2 are preset thresholds, 1≤t1≤5, 1≤t2≤5;

[0027] (3.2) If identical time series rules exist, increment the count of the corresponding time series rule by one; otherwise, increment r. i Add to set R k middle;

[0028] Step 4: For pulse group classes generated at different sorting cycles, calculate waveform similarity by comparing the timing rules contained in the pulse group classes to complete the pulse group class merging.

[0029] (4.1) For two pulse group classes C generated at different sorting beats m and C n The time series rules it contains are R m and R n , will R m and R n The timing rules in R are compared pairwise and compared to determine if they are the same: m any one of the timing rules and R n any one of the timing rules If the conditions described in formula (4) are met, then the timing rule r is considered to be true. p and r q same;

[0030]

[0031] where t3 and t4 are preset thresholds, 1 ≤ t3 ≤ 5, 1 ≤ t4 ≤ 5;

[0032] (4.2) Identify N eq as pulse group class C m and C n calculate the number of identical timing rules included in C m and C n for waveform similarity where N m is the number of timing rules included in R m and N n is the number of timing rules included in R n ;

[0033] (4.3) If the waveform similarity sim m,n is greater than the set threshold N1, 0 < N1 ≤ 0.8, then determine that C m and C n belong to the same radiation source, merge them, and add all the pulse groups included in C m and C n to the merged pulse group class C F .

Claims

1. A method for combining radiation source pulses based on inter-frame waveform similarity calculation, characterized in that: Step 1: Traverse the pulse groups intercepted within the current sorting cycle and generate pulse group classes based on pulse group parameters: For any two pulse groups with a fixed repetition period type, determine whether the two pulse groups overlap in the time domain. If there is no overlap, further determine whether the repetition period values ​​of the two are the same. If they are the same, determine that the two belong to the same class. If they are different, calculate the arrival time difference between the two and set a time threshold T1 to determine whether the two belong to the same class. Step 2: For the newly generated pulse group class C k Sort all pulse groups contained therein according to their arrival time and extract timing rules: Traverse C k For pulse groups, for adjacent pulse groups, calculate the arrival time difference between them; if the arrival time difference is within a preset time threshold T2, then generate a timing rule r = p pre →p suf , where p pre p is the repetition period value of the previous pulse group. suf This is the repetition period value of the next pulse group; Step 3: Statistically analyze and update pulse group class C k Timing rules in: For a newly generated timing rule r, match r with pulse group class C k The set of time-series rules R included k All time series rules in the set are compared pairwise to determine if they are identical. If identical time series rules are found, the count of the corresponding time series rule is incremented by one; otherwise, r is added to the set R. k middle; Step 4: Generate two pulse group classes C for different sorting cycles. m and C n The time series rules it contains are R m and R n Calculate C m and C n waveform similarity Where N eq For pulse group class C m and C n The number of identical timing rules contained, N m For R m The number of time-series rules included, N n For R n The number of timing rules included; If the waveform similarity is greater than the set threshold, then C is determined. m and C n Those belonging to the same radiation source are merged, and C is... m and C n All pulse groups included are added to the merged pulse group class C. F middle.

2. The radiation source pulse combination method based on inter-frame waveform similarity calculation according to claim 1, characterized in that: Step 1 further includes: (1.1) Traverse the pulse groups intercepted within the current sorting cycle. For any two pulse groups with a fixed repetition period, determine whether the two pulse groups overlap in the time domain. If the arrival time of the pulses contained in the two pulse groups satisfies the formula (1), then it is determined that the two do not overlap. WARRIOR f -TOA l >T d (1) Among them, TOA f TOA represents the arrival time of the first pulse in the next pulse group. l T represents the arrival time of the last pulse in the previous pulse group. d This indicates the preset time difference threshold; (1.2) If the two pulse groups do not overlap in the time domain, further determine whether the repetition period of the two pulse groups is the same; (1.3) If the repetition period of two pulse groups is the same, they are determined to belong to the same category. If they are different, the arrival time difference pgdTOA is calculated according to formula (2). If the arrival time difference is less than or equal to the set time threshold T1, they are determined to belong to the same category. Among them, TOA i TOA represents the arrival time of the first pulse in the next pulse group. j This indicates the arrival time of the first pulse in the previous pulse group.

3. The radiation source pulse combination method based on inter-frame waveform similarity calculation according to claim 1, characterized in that: Step 3 further includes: for the timing rule r = p pre →p suf and R k any one of the timing rules If the conditions described in formula (3) are met, then the timing rules r and r are considered to be true. p same; Where t1 and t2 are preset thresholds.

4. The radiation source pulse combination method based on inter-frame waveform similarity calculation according to claim 1, characterized in that: Step 4 also includes: (4.1) For two pulse group classes C generated at different sorting beats m and C n The time series rules it contains are R m and R n , will R m and R n The timing rules in R are compared pairwise to determine if they are the same. m any one of the timing rules and R n any one of the timing rules If the conditions described in formula (4) are met, then the timing rule r is considered to be true. p and r q same; Where t3 and t4 are preset thresholds, 1≤t3≤5, 1≤t4≤5; (4.2) N eq Identified as pulse group class C m and C n Calculate C for the number of identical timing rules included. m and C n waveform similarity Where N m For R m The number of time-series rules included, N n For R n The number of timing rules included; (4.3) If the waveform similarity sim m,n is greater than the set threshold N1, where 0 < N1 ≤ 0.8, then it is determined that C m and C n belong to the same radiation source, and they are merged. All pulse groups included in C m and C n are added to the merged pulse group class C F .

Images