CN113075637A - Airborne PD radar signal sorting method based on pulse descriptor data compression - Google Patents

Airborne PD radar signal sorting method based on pulse descriptor data compression Download PDF

Info

Publication number
CN113075637A
CN113075637A CN202110477112.6A CN202110477112A CN113075637A CN 113075637 A CN113075637 A CN 113075637A CN 202110477112 A CN202110477112 A CN 202110477112A CN 113075637 A CN113075637 A CN 113075637A
Authority
CN
China
Prior art keywords
pulse
sorting
pdw
data compression
pulse group
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Granted
Application number
CN202110477112.6A
Other languages
Chinese (zh)
Other versions
CN113075637B (en
Inventor
王聪
闫秋飞
宋新超
王澍
王星宇
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
723 Research Institute of CSIC
Original Assignee
723 Research Institute of CSIC
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 723 Research Institute of CSIC filed Critical 723 Research Institute of CSIC
Priority to CN202110477112.6A priority Critical patent/CN113075637B/en
Publication of CN113075637A publication Critical patent/CN113075637A/en
Application granted granted Critical
Publication of CN113075637B publication Critical patent/CN113075637B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
    • G01S7/00—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00
    • G01S7/02—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00 of systems according to group G01S13/00
    • G01S7/41—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00 of systems according to group G01S13/00 using analysis of echo signal for target characterisation; Target signature; Target cross-section
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
    • Y02A90/00—Technologies having an indirect contribution to adaptation to climate change
    • Y02A90/10—Information and communication technologies [ICT] supporting adaptation to climate change, e.g. for weather forecasting or climate simulation

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Radar, Positioning & Navigation (AREA)
  • Remote Sensing (AREA)
  • Radar Systems Or Details Thereof (AREA)

Abstract

The invention discloses an airborne PD radar signal sorting method based on pulse descriptor data compression, wherein in the initial stage, when a sorted target is detected to be an airborne PD radar signal, a data compression enabling identifier corresponding to a pre-sorting gate associated with the target is set to be effective; in subsequent processing, the signal pre-sorting functional module allocates pre-sorting channels for the input pulse description words by using a characteristic parameter matching method; and performing data compression processing on the pulse description words on the pre-sorting channels with the valid data compression enabling marks. The signal main sorting functional module extracts pulse group description words with compressed identifications from all pre-sorting channels, and classifies the extracted pulse group description word sequences by utilizing a nearest neighbor clustering algorithm; and sorting and searching the pulse group description words without compressed identifications by adopting a pulse sequence de-interlacing algorithm and a pulse group characteristic association algorithm respectively. The invention can accurately sort the airborne PD radar signals.

Description

Airborne PD radar signal sorting method based on pulse descriptor data compression
Technical Field
The invention belongs to the technical field of electronic reconnaissance system signal sorting, and particularly relates to an airborne PD radar signal sorting method based on pulse descriptor data compression.
Background
The electronic scout system signal sorting is to sort out Pulse PDW data belonging to different radar radiation sources from a sequence of staggered Pulse Description Words (PDWs) obtained by scout and interception. The Pulse descriptor PDW includes five parameters, i.e., Time of arrival (TOA), Direction of arrival (DOA), carrier Frequency (RF), Pulse Width (PW), and Pulse Amplitude (PA).
The signal sorting of the present electronic reconnaissance system mainly comprises a pre-sorting part and a main sorting part, wherein the pre-sorting part is used for sorting the input pulse sequence by using the characteristic parameters of orientation, carrier frequency, pulse width and the like in a pulse description word, and aims to dilute the density of a pulse data stream and reduce the complexity of subsequent processing; and the main sorting is to perform de-interleaving and sequence search processing on the pulse subsequences after the pre-sorting by using the pulse arrival time in the pulse description word. Because the airborne Pulse Doppler (PD) radar generally adopts the medium/high pulse repetition frequency (10-300 KHz), the signal density in the electromagnetic environment of a battlefield is multiplied, particularly when the signal density reaches million pulses per second (even higher), the electronic reconnaissance system is difficult to carry out real-time main sorting on all intercepted pulses, and when the number of the pulses stored in the pre-sorting gate reaches the upper limit and is not processed by the main sorting module in time, the pulse PDW data loss phenomenon can occur. In this case, the time relationship corresponding to each primary sorting result is discontinuous, and pulse groups with different frequencies and repetition frequency parameters are sorted into different results. These different sorting results can only be fused by means of target orientation information, empirical knowledge or databases. However, in practice, because the time relevance among the pulse groups is artificially destroyed, the electronic reconnaissance system serving as a non-cooperative party is difficult to acquire the prior information of feature quantities of enemy radar, and therefore the existing sorting algorithm is difficult to correctly sort the airborne PD radar signals.
Disclosure of Invention
The invention aims to provide an airborne PD radar signal sorting method based on pulse descriptor data compression so as to correctly sort airborne PD radar signals.
The technical solution for realizing the purpose of the invention is as follows: a pulse descriptor data compression-based airborne PD radar signal sorting method is provided with three modules of signal pre-sorting, signal main sorting and PD signal characteristic identification, and comprises the following specific steps:
step 1, in an initial stage, a signal preprocessing module does not perform data compression on pulse group description words in a pre-sorting channel, and if and only when a PD signal characteristic identification module detects that a sorted target is an airborne PD radar signal, a data compression enabling identifier corresponding to the pre-sorting channel associated with the target is set to be effective;
step 2, the signal pre-sorting module allocates a pre-sorting channel for the input pulse description word by using a characteristic parameter matching method, and judges whether a data compression enabling identifier corresponding to the pre-sorting channel is effective or not: if yes, entering step 3; otherwise, entering step 4;
step 3, carrying out data compression processing on the pulse description word according to a space-time trajectory data compression principle on a pre-sorting channel with the effective data compression enabling identifier to form a pulse group description word with a compression identifier;
step 4, judging whether all the pulse description words to be processed in the data cache region are processed completely, if all the pulse description words are processed completely, starting to execute main sorting, namely executing step 5, otherwise, returning to step 2;
step 5, extracting pulse group description words with compressed identifications from all pre-sorting channels by a signal main sorting module, and classifying the extracted pulse group description word sequences by utilizing a nearest neighbor clustering algorithm according to the azimuth information of the pulse group description words;
step 6, respectively carrying out feature association processing on the pulse group description words in each category;
and 7, performing pulse de-interleaving and sequence search processing on the pulse group description words without the compression identifiers in each pre-sorting channel by using a pulse sequence de-interleaving algorithm.
Compared with the prior art, the invention has the following remarkable advantages: (1) under the condition of not losing important information of a target radiation source, data compression is carried out on pulse description words in the pre-sorting channel according to the space-time trajectory data compression principle to form pulse group description words with compressed identifiers, so that the loss probability of pulses during signal pre-sorting is reduced, and the sorting capacity of the electronic reconnaissance system for medium/high-repetition frequency PD radar signals is improved; (2) in the signal main sorting stage, a classical pulse sequence de-interlacing algorithm and a pulse group characteristic correlation algorithm are adopted to respectively sort and sequence search pulse group description words without compression identification and with compression identification, and a more effective discrimination basis is provided for an airborne pulse Doppler radar identification technology.
Drawings
Fig. 1 is a general block diagram of airborne PD radar signal sorting.
Fig. 2 is a schematic diagram of the structure of a pulse group descriptor word.
FIG. 3 is a schematic diagram of pulse data compression within a pre-sort channel.
Detailed Description
The invention relates to a pulse descriptor data compression-based airborne PD radar signal sorting method, which is provided with three modules of signal pre-sorting, signal main sorting and PD signal characteristic identification, and comprises the following specific steps:
step 1, in an initial stage, a signal preprocessing module does not perform data compression on pulse group description words in a pre-sorting channel, and if and only when a PD signal characteristic identification module detects that a sorted target is an airborne PD radar signal, a data compression enabling identifier corresponding to the pre-sorting channel associated with the target is set to be effective;
step 2, the signal pre-sorting module allocates a pre-sorting channel for the input pulse description word by using a characteristic parameter matching method, and judges whether a data compression enabling identifier corresponding to the pre-sorting channel is effective or not: if yes, entering step 3; otherwise, entering step 4;
step 3, carrying out data compression processing on the pulse description word according to a space-time trajectory data compression principle on a pre-sorting channel with the effective data compression enabling identifier to form a pulse group description word with a compression identifier;
step 4, judging whether all the pulse description words to be processed in the data cache region are processed completely, if all the pulse description words are processed completely, starting to execute main sorting, namely executing step 5, otherwise, returning to step 2;
step 5, extracting pulse group description words with compressed identifications from all pre-sorting channels by a signal main sorting module, and classifying the extracted pulse group description word sequences by utilizing a nearest neighbor clustering algorithm according to the azimuth information of the pulse group description words;
step 6, respectively carrying out feature association processing on the pulse group description words in each category;
and 7, performing pulse de-interleaving and sequence search processing on the pulse group description words without the compression identifiers in each pre-sorting channel by using a pulse sequence de-interleaving algorithm.
As a specific example, the signal pre-sorting module in step 2 allocates pre-sorting channels to the input pulse description words by using a characteristic parameter matching method, which is as follows:
(2.1) the signal pre-sorting function module reads a pulse description word to be processed from the data processing buffer area, and the pulse description word is recorded as pdwi＝{toai，doai，rfi，pwi，paiWhere i denotes the read pulse description word number, toai、doai、rfi、pwi、paiRespectively representing the arrival time, the direction, the carrier frequency, the pulse width and the pulse amplitude value of the pulse;
(2.2) calculating the pulse descriptor pdwiThe similarity between the pre-divided gates and the established pre-divided gates is set as the feature vector set corresponding to the established pre-divided gates
Figure BDA0003047465450000034
J is more than 0 and less than or equal to CNum, and the CNum is the number of the established pre-sorting channels; xjFor the feature vector corresponding to the jth pre-branch gate,
Figure BDA0003047465450000031
wherein
Figure BDA0003047465450000032
Figure BDA0003047465450000033
Respectively representing the position, carrier frequency and pulse width value corresponding to the jth pre-sorting gate;
pdwiand XjSimilarity u ofjThe calculation method comprises the following steps:
uj＝α1×u_doaj+α2×u_rfj+α3×u_pwj
wherein alpha is1、α2、α3Respectively, are the weight coefficients of the image data,
Figure BDA0003047465450000041
Figure BDA0003047465450000042
respectively representing the similarity of the azimuth, the carrier frequency and the pulse width;
the function Φ is defined as follows:
Figure BDA0003047465450000043
where σ is the measurement error of the parameter and y represents the pulse descriptor pdwiThe corresponding characteristic parameter(s) in (1),
Figure BDA0003047465450000044
representing the characteristic parameter corresponding to the jth pre-sorting gate; all the finally obtained similarities were recorded as sequences
Figure BDA0003047465450000048
0＜j≤CNum；
(2.3) calculating the maximum value of similarity
Figure BDA0003047465450000045
Marking the pre-branch gate channel corresponding to the maximum value as BestCH, and marking u asmaxComparing with a decision threshold THR _ u: such asFruit umax< THR _ u, then pdw is illustratediThe pre-sorting channels are not matched, and the step (2.4) is executed; if u ismax≧ THR _ u, then caption pdwiMatching with a pre-branch channel BestCH, and executing the step (2.5);
(2.4) at pdwiIn order to establish a new pre-branch gate NewCH according to the method, the characteristic vector corresponding to the pre-branch gate is as follows:
Figure BDA0003047465450000046
building a first burst descriptor, GPDW, in the pre-branch channel NewCH1＝{PDW1，EXT1，FG1In which PDW1For pulse description word fields, EXT1For additional fields, FG1Is an identification bit word; will pulse description word pdwiStored in PDW field, i.e. PDW1＝pdwi＝{toai，doai，rfi，pwi，pai}, additional field EXT1The values of (A) are as follows:
Figure BDA0003047465450000047
the flag bit field is set to "invalid", i.e.: let FG10, indicating that the pulse group description word has no data compression; number of pulse groups PNum in a pre-sort channelBestCHStep 4 is executed as 1;
(2.5) correcting the eigenvector corresponding to the BestCH of the pre-branch channel
Figure BDA0003047465450000051
Namely:
Figure BDA0003047465450000052
wherein, PNumBestCHThe number of pulse groups in the pre-sorting channel; then step 3 is performed.
As a specific example, in step 3, the pre-sorting channel that enables the data compression identification to be valid performs data compression processing on the pulse description word according to the space-time trajectory data compression principle to form a pulse group description word with a compression identification, which specifically is:
(3.1) let p ═ PNumBestCHP represents the index number of the pulse group description word in the pre-branch strobe BestCH;
(3.2) calculation pdwiAnd GPDWpThe difference in time between Δ T, i.e. Δ T ═ toai-toaepWherein toaiIs pdwiTime of arrival of pulse in (toae)pIs GPDWpA burst termination time in the additional field;
(3.3) determination of GPDWpWhether the flag field of (G) is valid, i.e. determining FGpWhether or not 1 holds: if so, performing step (3.4); otherwise, executing the step (3.6);
(3.4) order
Figure BDA0003047465450000053
Wherein
Figure BDA0003047465450000054
For the lower rounding operator, PRIlThe pulse repetition interval value used by the PD radar associated with the pre-branch gate is more than or equal to 1 and less than or equal to PRIQ, and the PRIQ is the number of the pulse repetition interval values; finding the sequence εl1 ≦ l ≦ the minimum value ε of PRIQ, i.e
Figure BDA0003047465450000055
The corresponding pulse interval is noted as PRIminComparing epsilon with a decision threshold eta, if epsilon > eta, then it is stated that pdw cannot be usediData compression to GPDW in (1)pStep (3.8) is performed; otherwise, the description can refer to pdwiData compression to GPDW in (1)pIn the method, the data compression method is change GPDWpInformation of the additional fields, namely:
Figure BDA0003047465450000061
(3.5) applying GPDWpIs set to valid, i.e.: let FGpStep 4 is then performed;
(3.6) order
Figure BDA0003047465450000062
Wherein
Figure BDA0003047465450000063
For the lower rounding operator, calculate ε ═ Δ T-a × pripComparing epsilon with a decision threshold eta: if ε > η or a is 0, then the description will not refer to pdwiData compression to GPDW in (1)pStep (3.8) is performed; otherwise, the description can refer to pdwiData compression to GPDW in (1)pIn the method, the data compression method is change GPDWpInformation of the additional fields, namely:
Figure BDA0003047465450000064
(3.7) application of GPDWpIs set to "valid", i.e.: let FGpStep 4 is then performed;
(3.8) let p be p-1, if p > 0, return to step (3.3); otherwise, pdw cannot be explainediCompressing the pulse group description word into the existing pulse group description word in the pre-branch channel BestCH, and executing the step (3.9);
(3.9) adding a new pulse group description word in the presorted gate BestCH, and changing the number of pulse groups in the presorted gate into PNumBestCH＝PNumBestCH+1, denote the newly added burst descriptor as GPDWk＝{PDWk，EXTk，FGkWhere k is PNumBestCH(ii) a Setting GPDWkThe values of the fields in (1) are as follows: will pulse description word pdwiThe information in (1) is copied to the PDW field, i.e. PDWk＝pdwi＝{toak，doak，rfki，pwk，pak}; setting the flag bit fieldAnd (4) setting to be invalid, namely: let FGk0, no data compression of the burst descriptor, and an additional field EXTkThe initialization information of (1) is as follows:
Figure BDA0003047465450000071
as a specific example, in step 6, the pulse group description words in each category are respectively subjected to feature association processing, and a category obtained by classifying a pulse group description word sequence is first marked as { ClustersS is more than or equal to 1 and less than or equal to ClusterNum, the ClusterNum is the number of categories, and the categories are sequentially matched with the { ClustersPerforming feature association processing on the pulse group description words in the description, specifically as follows:
(6.1) setting the class ClustersThe number of pulse group description words contained in the above-mentioned description words is GNum, and the class Cluster is definedsThe pulse group description words in (1) are reordered according to the pulse group starting time, and the sequence of the ordered pulse group description words is marked as { gpdwnN is more than or equal to 1 and less than or equal to GNum;
(6.2) let n11, wherein n1For a sequence of pulse groups { gpdwnA subscript of };
(6.3) let n2＝n1+1, wherein n2For a sequence of pulse groups { gpdwnA subscript of };
(6.4) according to
Figure BDA0003047465450000072
Corresponding pulse group start time
Figure BDA0003047465450000073
And a termination time
Figure BDA0003047465450000074
And
Figure BDA0003047465450000075
corresponding pulse group start time
Figure BDA0003047465450000076
And a termination time
Figure BDA0003047465450000077
Judgment of
Figure BDA0003047465450000078
And
Figure BDA0003047465450000079
whether staggered in time: if it is not
Figure BDA00030474654500000710
Or
Figure BDA00030474654500000711
Then the two pulse groups are not staggered in time, and step (6.5) is executed; otherwise, the two pulse groups are staggered in time, and step (6.6) is executed;
(6.5) let n1＝n2If n is1< GNum, then return to step (6.3); otherwise, the pulse group description word association is finished, and step (6.7) is executed;
(6.6) let n2＝n2+1, if n2GNum is less than or equal to, then the step (6.4) is returned; otherwise, the pulse group description word association is finished, and step (6.7) is executed;
(6.7) mixing with
Figure BDA00030474654500000712
And separating the successfully associated pulse group description words from the classes, and then associating the rest pulse group description words in the classes in the same way until all the pulse group description words in the classes are completely classified.
The present invention will be described in further detail with reference to FIGS. 1 to 3 and specific examples.
Examples
The invention discloses an airborne PD radar signal sorting method based on pulse descriptor data compression. Fig. 1 is a general block diagram of the implementation of the method of the present invention. As shown in FIG. 1, the method of the invention mainly comprises three functional modules of signal pre-sorting, signal main sorting and PD signal characteristic identification. In the initial stage, the signal preprocessing function module does not perform data compression processing on the pulse description word, and if and only if the PD signal characteristic identification module detects that the sorted target is an airborne PD radar signal, the data compression enable identifier corresponding to the pre-sorting gate associated with the target is set to "valid". In the subsequent processing, a signal pre-sorting functional module firstly allocates pre-sorting channels for the input pulse description words by using a characteristic parameter matching method; and then, carrying out data compression processing on the pulse description word according to a space-time trajectory data compression principle on the pre-branch gate with the data compression enabling identifier being 'effective' to form the pulse group description word with the compression identifier. The signal main sorting functional module extracts pulse group description words with 'compressed identification' from all pre-sorting channels, and classifies the extracted pulse group description word sequences by utilizing a nearest neighbor clustering algorithm according to the azimuth information of the pulse group description words; carrying out characteristic association processing on the pulse group description words in each category; and respectively carrying out pulse de-interleaving and sequence search processing on each pre-sorting channel by using a classical pulse sequence de-interleaving algorithm on pulse group description words without 'compression identification' in each pre-sorting channel.
Step 1, a PD signal characteristic identification function module identifies the airborne PD radar signal characteristics of the selected target, and when the target is found to be the airborne PD radar signal, a data compression enabling identifier corresponding to a pre-sorting gate associated with the target is set to be effective.
Step 2, the signal pre-sorting function module reads a pulse description word to be processed from the data processing buffer area, and the pulse description word is recorded as pdwi＝{toai，doai，rfi，pwi，paiWhere i denotes the read pulse description word number, toai、doai、rfi、pwi、paiRespectively representing the arrival time, the azimuth, the carrier frequency, the pulse width and the pulse amplitude value of the pulse.
Step 3, calculating pulse description word pdwiAnd has already establishedThe similarity between the pre-branch gates does not set the feature vector set corresponding to the pre-branch gates established as
Figure BDA0003047465450000081
(where j is greater than 0 and less than or equal to CNum, which is the number of pre-sort channels already established), XjFor the feature vector corresponding to the jth pre-branch gate,
Figure BDA0003047465450000082
wherein the content of the first and second substances,
Figure BDA0003047465450000083
respectively showing the corresponding position, carrier frequency and pulse width value of the jth pre-sorting gate. pdwiAnd XjSimilarity u ofjThe calculation method comprises the following steps:
uj＝α1×u_doaj+α2×u_rfj+a3×u_pwj
wherein alpha is1、α2、α3Respectively, are the weight coefficients of the image data,
Figure BDA0003047465450000091
Figure BDA0003047465450000092
respectively representing the similarity of the azimuth, the carrier frequency and the pulse width. The function Φ is defined as follows:
Figure BDA0003047465450000093
where σ is the measurement error of the parameter and y represents the pulse descriptor pdwiThe corresponding characteristic parameter(s) in (1),
Figure BDA0003047465450000094
and representing the characteristic parameters corresponding to the jth pre-sorting channel. All the finally obtained similarities were recorded as sequences
Figure BDA0003047465450000095
(0＜j≤CNum)。
Step 4, calculating the maximum value of the similarity
Figure BDA0003047465450000096
And recording the pre-branch gate corresponding to the maximum value as BestCH. Will umaxComparing with a decision threshold THR _ u: if u ismax< THR _ u, then pdw is illustratediStep 5 is executed if the signal is not matched with the pre-sorting gate; if u ismax≧ THR _ u, then caption pdwiMatching with a pre-branch gate BestCH, and executing a step 6;
step 5, pdwiIn order to establish a new pre-branch gate NewCH according to the method, the characteristic vector corresponding to the pre-branch gate is as follows:
Figure BDA0003047465450000097
building a first burst descriptor, GPDW, in the pre-branch channel NewCH1＝{PDW1，EXT1，FG1In which PDW1For pulse description word fields, EXT1For additional fields, FG1For identifying the bit words, the details of the pulse group description word are shown in fig. 2. Will pulse description word pdwiStored in PDW field, i.e. PDW1＝pdwi＝{toai，doai，rfi，pwi，paiEXT additional field1The values of (A) are as follows:
Figure BDA0003047465450000098
the flag bit field is set to "invalid", i.e.: let FG10, indicating that the burst descriptor has no data compression. Number of pulse groups PNum in a pre-sort channelBestCHStep 9 is performed as 1.
Step 6, correcting the eigenvector corresponding to the pre-branch gate BestCH
Figure BDA0003047465450000101
Namely:
Figure BDA0003047465450000102
wherein, PNumBestCHThe number of pulse groups in the pre-sort channel.
And 7, judging whether the data compression enabling identification corresponding to the pre-branch gate BestCH is effective or not. If the data compression enable flag for the pre-sort channel is "valid," then step 8 is performed; otherwise step 9 is performed.
Step 8, pulse description words pdw are compressed according to the space-time trajectory data compression principleiCarrying out data compression treatment, which comprises the following specific steps:
step 8-1, let p ═ PNumBestCHWherein p represents the index number of the pulse group description word in the pre-branch strobe BestCH;
step 8-2, calculate pdwiAnd GPDWpThe difference in time between Δ T, i.e.: Δ T ═ toai-toaepWherein toaiIs pdwiTime of arrival of pulse in (toae)pIs GPDWpThe pulse group end time in the additional field.
Step 8-3, judging GPDWpWhether the flag field of (G) is "valid", that is, the FG is judgedpIs 1 true? If true, then step 8-4 is performed; otherwise, executing step 8-6;
step 8-4, order
Figure BDA0003047465450000103
(wherein
Figure BDA0003047465450000104
For the lower rounding operator, PRIlThe pulse repetition interval value used by the PD radar associated with the pre-gated channel, l is more than or equal to 1 and less than or equal to PRIQ, the number of PRIQ pulse repetition interval values), finding out the sequence { epsilon ∈lThe minimum value ε of (1. ltoreq. l.ltoreq.PRIQ), i.e.
Figure BDA0003047465450000105
(the corresponding pulse interval is denoted as PRImin) Comparing epsilon with a decision threshold eta, if epsilon > eta, then it is stated that pdw cannot be usediData compression to GPDW in (1)pStep 8-8 is performed; otherwise, the description can refer to pdwiData compression to GPDW in (1)pIn the method, the data compression method is change GPDWpInformation of the additional fields, namely:
Figure BDA0003047465450000111
step 8-5, GPDWpIs set to "valid", i.e.: let FGpStep 9 is executed as 1;
step 8-6, order
Figure BDA0003047465450000112
(wherein
Figure BDA0003047465450000113
For the lower rounding operator), calculate ∈ ═ Δ T-a × pripComparing epsilon with a decision threshold eta: if ε > η or a is 0, then the description will not refer to pdwiData compression to GPDW in (1)pStep 8-8 is performed; otherwise, the description can refer to pdwiData compression to GPDW in (1)pIn the method, the data compression method is change GPDWpInformation of the additional fields, namely:
Figure BDA0003047465450000114
step 8-7, GPDWpIs set to "valid", i.e.: let FGpStep 9 is executed as 1;
step 8-8, making p equal to p-1, and if p is greater than 0, returning to step 8-3; otherwise, pdw cannot be explainediCompressing the pulse group description word into the existing pulse group description word in the pre-branch gate BestCH, and executing the step 8-9;
step 8-9 in advanceA new pulse group description word is added in the BestCH of the sorting channel, and the number of the pulse groups in the pre-sorting channel is changed into PNumBestCH＝PNumBestCH+1, denote the newly added burst descriptor as GPDWk＝{PDWk，EXTk，FGkWhere k is PNumBestCH(ii) a Setting GPDWkThe values of the fields in (1) are as follows: will pulse description word pdwiThe information in (1) is copied to the PDW field, i.e. PDWk＝pdwi＝{toak，doak，rfki，pwk，pak}; the flag bit field is set to "invalid", i.e.: let FGk0; indicating that the burst description word has no data compression. Additional field EXTkThe initialization information of (1) is as follows:
Figure BDA0003047465450000121
and 9, judging whether all the pulse description words to be processed in the data cache region are processed completely, if all the pulse description words are processed completely, starting to execute main sorting, namely executing the step 10, and otherwise, returning to the step 2.
Step 10, extracting pulse group description words with 'compressed identifications' from all pre-sorting channels, classifying the extracted pulse group description word sequences by utilizing a nearest neighbor clustering algorithm according to the azimuth information of the pulse group description words, and marking the obtained classes as { ClustersAnd s is more than or equal to 1 and less than or equal to ClusterNum, wherein the ClusterNum is the number of categories.
Step 11, sequentially matching the categories { ClustersThe pulse group description words in the (s is more than or equal to 1 and less than or equal to ClusterNum) are subjected to characteristic association processing, and the category Cluster is used belowsFor example, a method for performing feature association processing on a pulse group descriptor is introduced, and the method specifically includes the following steps:
step 11-1, setting no class ClustersThe number of pulse group description words contained in the above-mentioned description words is GNum, and the class Cluster is definedsThe pulse group description words in (1) are reordered according to the pulse group starting time, and the sequence of the ordered pulse group description words is marked as { gpdwn} of whichWherein n is more than or equal to 1 and less than or equal to GNum.
Step 11-2, let n11, wherein n1For a sequence of pulse groups { gpdwnThe subscript of.
Step 11-3, let n2＝n1+1, wherein n2For a sequence of pulse groups { gpdwnThe subscript of.
Step 11-4, according to
Figure BDA0003047465450000122
Corresponding pulse group start time
Figure BDA0003047465450000123
And a termination time
Figure BDA0003047465450000124
And
Figure BDA0003047465450000125
corresponding pulse group start time
Figure BDA0003047465450000126
And a termination time
Figure BDA0003047465450000127
Judgment of
Figure BDA0003047465450000128
And
Figure BDA0003047465450000129
is it staggered in time? If it is not
Figure BDA00030474654500001210
Or
Figure BDA00030474654500001211
Then the two pulse groups are not interleaved in time and step 11-5 is performed; otherwise, the two pulse groups are said to be interleaved in time, and step 11-6 is performed.
Step 11-5, let n1＝n2If n is1< GNum, then return to step 11-3; otherwise, the pulse group description word association is finished, and step 11-7 is executed;
step 11-6, let n2＝n2+1, if n2GNum is less than or equal to GNum, then the step 11-4 is returned; otherwise, the pulse group description word association is finished, and step 11-7 is executed;
step 11-7, mixing
Figure BDA0003047465450000131
And separating the successfully associated pulse group description words from the classes, and then associating the rest pulse group description words in the classes in the same way until all the pulse group description words in the classes are completely classified.
And step 12, performing pulse de-interleaving and sequence searching processing on the pulse group description words without the compression identifiers in each pre-sorting channel, wherein the pulse group description words do not meet the characteristics of the airborne PD radar, and therefore the pulse group description words can be processed by using a classical pulse sequence de-interleaving algorithm, which is not described in detail here.
The invention reduces the loss probability of the pulse during signal pre-sorting, improves the sorting capability of the electronic reconnaissance system for the PD radar signals with medium/high repetition frequency, and can provide more effective discrimination basis for the identification technology of the airborne pulse Doppler radar.

Claims (4)

1. The airborne PD radar signal sorting method based on pulse descriptor data compression is characterized by comprising three modules of signal pre-sorting, signal main sorting and PD signal characteristic identification, and comprises the following specific steps:
step 1, in an initial stage, a signal preprocessing module does not perform data compression on pulse group description words in a pre-sorting channel, and if and only when a PD signal characteristic identification module detects that a sorted target is an airborne PD radar signal, a data compression enabling identifier corresponding to the pre-sorting channel associated with the target is set to be effective;
step 2, the signal pre-sorting module allocates a pre-sorting channel for the input pulse description word by using a characteristic parameter matching method, and judges whether a data compression enabling identifier corresponding to the pre-sorting channel is effective or not: if yes, entering step 3; otherwise, entering step 4;
step 3, carrying out data compression processing on the pulse description word according to a space-time trajectory data compression principle on a pre-sorting channel with the effective data compression enabling identifier to form a pulse group description word with a compression identifier;
step 4, judging whether all the pulse description words to be processed in the data cache region are processed completely, if all the pulse description words are processed completely, starting to execute main sorting, namely executing step 5, otherwise, returning to step 2;
step 5, extracting pulse group description words with compressed identifications from all pre-sorting channels by a signal main sorting module, and classifying the extracted pulse group description word sequences by utilizing a nearest neighbor clustering algorithm according to the azimuth information of the pulse group description words;
step 6, respectively carrying out feature association processing on the pulse group description words in each category;
and 7, performing pulse de-interleaving and sequence search processing on the pulse group description words without the compression identifiers in each pre-sorting channel by using a pulse sequence de-interleaving algorithm.
2. The method for sorting the airborne PD radar signals based on pulse descriptor data compression according to claim 1, characterized in that, the signal pre-sorting module in step 2 allocates pre-sorting channels to the input pulse descriptors using a characteristic parameter matching method, specifically as follows:
(2.1) the signal pre-sorting function module reads a pulse description word to be processed from the data processing buffer area, and the pulse description word is recorded as pdwi＝{toai，doai，rfi，pwi，paiWhere i denotes the read pulse description word number, toai、doai、rfi、pwi、paiRespectively representing the arrival time, the direction, the carrier frequency, the pulse width and the pulse amplitude value of the pulse;
(2.2) calculating the pulse descriptor pdwiAnd has already establishedThe similarity between the pre-branch gates is set as the feature vector set corresponding to the pre-branch gates to be { X }jJ is more than 0 and less than or equal to CNum, and the CNum is the number of the established pre-sorting channels; xjFor the feature vector corresponding to the jth pre-branch gate,
Figure FDA0003047465440000011
wherein
Figure FDA0003047465440000012
Figure FDA0003047465440000013
Respectively representing the position, carrier frequency and pulse width value corresponding to the jth pre-sorting gate;
pdwiand XjSimilarity u ofjThe calculation method comprises the following steps:
uj＝α1×u_doaj+α2×u_rfj+α3×u_pwj
wherein alpha is1、α2、α3Respectively, are the weight coefficients of the image data,
Figure FDA0003047465440000021
Figure FDA0003047465440000022
respectively representing the similarity of the azimuth, the carrier frequency and the pulse width;
the function Φ is defined as follows:
Figure FDA0003047465440000023
where σ is the measurement error of the parameter and y represents the pulse descriptor pdwiThe corresponding characteristic parameter(s) in (1),
Figure FDA0003047465440000024
represents the jth pre-partitionCharacteristic parameters corresponding to the gating channels; all the finally obtained similarities are recorded as a sequence { u }j}，0＜j≤CNum；
(2.3) calculating the maximum value of similarity
Figure FDA0003047465440000025
Marking the pre-branch gate channel corresponding to the maximum value as BestCH, and marking u asmaxComparing with a decision threshold THR _ u: if u ismax< THR _ u, then pdw is illustratediThe pre-sorting channels are not matched, and the step (2.4) is executed; if u ismax≧ THR _ u, then caption pdwiMatching with a pre-branch channel BestCH, and executing the step (2.5);
(2.4) at pdwiIn order to establish a new pre-branch gate NewCH according to the method, the characteristic vector corresponding to the pre-branch gate is as follows:
Figure FDA0003047465440000026
building a first burst descriptor, GPDW, in the pre-branch channel NewCH1＝{PDW1，EXT1，FG1In which PDW1For pulse description word fields, EXT1For additional fields, FG1Is an identification bit word; will pulse description word pdwiStored in PDW field, i.e. PDW1＝pdwi＝{toai，doai，rfi，pwi，pai}, additional field EXT1The values of (A) are as follows:
Figure FDA0003047465440000031
the flag bit field is set to "invalid", i.e.: let FG10, indicating that the pulse group description word has no data compression; number of pulse groups PNum in a pre-sort channelBestCHStep 4 is executed as 1;
(2.5) correcting the eigenvector corresponding to the BestCH of the pre-branch channel
Figure FDA0003047465440000032
Namely:
Figure FDA0003047465440000033
wherein, PNumBestCHThe number of pulse groups in the pre-sorting channel; then step 3 is performed.
3. The method for sorting the airborne PD radar signals according to claim 2, characterized in that, in step 3, the pre-sorting channel that enables the data compression identification to be valid performs data compression processing on the pulse description words according to the space-time trajectory data compression principle to form pulse group description words with compressed identification, specifically:
(3.1) let p ═ PNumBestCHP represents the index number of the pulse group description word in the pre-branch strobe BestCH;
(3.2) calculation pdwiAnd GPDWpThe difference in time between Δ T, i.e. Δ T ═ toai-toaepWherein toaiIs pdwiTime of arrival of pulse in (toae)pIs GPDWpA burst termination time in the additional field;
(3.3) determination of GPDWpWhether the flag field of (G) is valid, i.e. determining FGpWhether or not 1 holds: if so, performing step (3.4); otherwise, executing the step (3.6);
(3.4) order
Figure FDA0003047465440000034
Wherein
Figure FDA0003047465440000035
For the lower rounding operator, PRIlThe pulse repetition interval value used by the PD radar associated with the pre-branch gate is more than or equal to 1 and less than or equal to PRIQ, and the PRIQ is the number of the pulse repetition interval values; finding the sequence εl1 ≦ l ≦ the minimum value ε of PRIQ, i.e
Figure FDA0003047465440000041
The corresponding pulse interval is noted as PRIminComparing epsilon with a decision threshold eta, if epsilon > eta, then it is stated that pdw cannot be usediData compression to GPDW in (1)pStep (3.8) is performed; otherwise, the description can refer to pdwiData compression to GPDW in (1)pIn the method, the data compression method is change GPDWpInformation of the additional fields, namely:
Figure FDA0003047465440000042
(3.5) applying GPDWpIs set to valid, i.e.: let FGpStep 4 is then performed;
(3.6) order
Figure FDA0003047465440000043
Wherein
Figure FDA0003047465440000044
For the lower rounding operator, calculate ε ═ Δ T-a × pripComparing epsilon with a decision threshold eta: if ε > η or a is 0, then the description will not refer to pdwiData compression to GPDW in (1)pStep (3.8) is performed; otherwise, the description can refer to pdwiData compression to GPDW in (1)pIn the method, the data compression method is change GPDWpInformation of the additional fields, namely:
Figure FDA0003047465440000045
(3.7) application of GPDWpIs set to "valid", i.e.: let FGpStep 4 is then performed;
(3.8) let p be p-1, if p > 0, return to step (3.3); otherwise, pdw cannot be explainediCompressing the pulse group description word into the existing pulse group description word in the pre-branch channel BestCH, and executing the step (3.9);
(3.9) adding a new pulse group description word in the presorted gate BestCH, and changing the number of pulse groups in the presorted gate into PNumBestCH＝PNumBestCH+1, denote the newly added burst descriptor as GPDWk＝{PDWk，EXTk，FGkWhere k is PNumBestCH(ii) a Setting GPDWkThe values of the fields in (1) are as follows: will pulse description word pdwiThe information in (1) is copied to the PDW field, i.e. PDWk＝pdwi＝{toak，doak，rfki，pwk，pak}; the flag bit field is set to invalid, i.e.: let FGk0, no data compression of the burst descriptor, and an additional field EXTkThe initialization information of (1) is as follows:
Figure FDA0003047465440000051
4. the method for sorting airborne PD radar signals based on pulse descriptor data compression as claimed in claim 3, characterized in that, in step 6, the pulse group descriptor in each category is processed by feature association, and the category obtained by first classifying the pulse group descriptor sequence is marked as { ClustersS is more than or equal to 1 and less than or equal to ClusterNum, the ClusterNum is the number of categories, and the categories are sequentially matched with the { ClustersPerforming feature association processing on the pulse group description words in the description, specifically as follows:
(6.1) setting the class ClustersThe number of pulse group description words contained in the above-mentioned description words is GNum, and the class Cluster is definedsThe pulse group description words in (1) are reordered according to the pulse group starting time, and the sequence of the ordered pulse group description words is marked as { gpdwnN is more than or equal to 1 and less than or equal to GNum;
(6.2) let n11, wherein n1For a sequence of pulse groups { gpdwnA subscript of };
(6.3) let n2＝n1+1, wherein n2For a sequence of pulse groups { gpdwnA subscript of };
(6.4) according to
Figure FDA0003047465440000052
Corresponding pulse group start time
Figure FDA0003047465440000053
And a termination time
Figure FDA0003047465440000054
And
Figure FDA0003047465440000055
corresponding pulse group start time
Figure FDA0003047465440000056
And a termination time
Figure FDA0003047465440000057
Judgment of
Figure FDA0003047465440000058
And
Figure FDA0003047465440000059
whether staggered in time: if it is not
Figure FDA0003047465440000061
Or
Figure FDA0003047465440000062
Then the two pulse groups are not staggered in time, and step (6.5) is executed; otherwise, the two pulse groups are staggered in time, and step (6.6) is executed;
(6.5) let n1＝n2If n is1< GNum, then return to step (6.3); otherwise, this isFinishing the association of the secondary pulse group description words, and executing the step (6.7);
(6.6) let n2＝n2+1, if n2GNum is less than or equal to, then the step (6.4) is returned; otherwise, the pulse group description word association is finished, and step (6.7) is executed;
(6.7) mixing with
Figure FDA0003047465440000063
And separating the successfully associated pulse group description words from the classes, and then associating the rest pulse group description words in the classes in the same way until all the pulse group description words in the classes are completely classified.
CN202110477112.6A 2021-04-29 2021-04-29 Airborne PD radar signal sorting method based on pulse descriptor data compression Active CN113075637B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202110477112.6A CN113075637B (en) 2021-04-29 2021-04-29 Airborne PD radar signal sorting method based on pulse descriptor data compression

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202110477112.6A CN113075637B (en) 2021-04-29 2021-04-29 Airborne PD radar signal sorting method based on pulse descriptor data compression

Publications (2)

Publication Number Publication Date
CN113075637A true CN113075637A (en) 2021-07-06
CN113075637B CN113075637B (en) 2022-06-07

Family

ID=76616071

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202110477112.6A Active CN113075637B (en) 2021-04-29 2021-04-29 Airborne PD radar signal sorting method based on pulse descriptor data compression

Country Status (1)

Country Link
CN (1) CN113075637B (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116821097A (en) * 2023-07-05 2023-09-29 中国人民解放军92728部队 Full pulse data processing method

Citations (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107576942A (en) * 2017-07-27 2018-01-12 中国船舶重工集团公司第七二四研究所 A kind of radiation source scan period real-time estimation method based on signal cluster
CN107656245A (en) * 2017-08-22 2018-02-02 哈尔滨工程大学 A kind of method being applied to information fusion in radar signal sorting
CN108549061A (en) * 2018-04-24 2018-09-18 扬州健行电子科技有限公司 A kind of clustering method of signal
CN109613486A (en) * 2018-12-03 2019-04-12 中国人民解放军空军工程大学 A kind of Radar Signal Sorting Method based on core cluster support vector clustering
CN110895331A (en) * 2019-08-21 2020-03-20 常熟理工学院 Pulse Doppler radar target sparse detection method based on structured observation matrix
CN111476288A (en) * 2020-04-03 2020-07-31 中国人民解放军海军航空大学 Intelligent perception method for cognitive sensor network to electromagnetic behaviors with unknown threats
CN111693944A (en) * 2020-06-18 2020-09-22 上海志良电子科技有限公司 Radar active interference signal parameter extraction and interference pattern identification method and device
CN111722188A (en) * 2020-05-18 2020-09-29 中国人民解放军63892部队 PRI (pulse repetition index) conversion radar signal sorting method based on STFT (space time Fourier transform) pre-sorting
CN112526455A (en) * 2020-12-25 2021-03-19 扬州健行电子科技有限公司 Method for splitting radar-spread-resistant signal into fixed and jittered signals

Patent Citations (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107576942A (en) * 2017-07-27 2018-01-12 中国船舶重工集团公司第七二四研究所 A kind of radiation source scan period real-time estimation method based on signal cluster
CN107656245A (en) * 2017-08-22 2018-02-02 哈尔滨工程大学 A kind of method being applied to information fusion in radar signal sorting
CN108549061A (en) * 2018-04-24 2018-09-18 扬州健行电子科技有限公司 A kind of clustering method of signal
CN109613486A (en) * 2018-12-03 2019-04-12 中国人民解放军空军工程大学 A kind of Radar Signal Sorting Method based on core cluster support vector clustering
CN110895331A (en) * 2019-08-21 2020-03-20 常熟理工学院 Pulse Doppler radar target sparse detection method based on structured observation matrix
CN111476288A (en) * 2020-04-03 2020-07-31 中国人民解放军海军航空大学 Intelligent perception method for cognitive sensor network to electromagnetic behaviors with unknown threats
CN111722188A (en) * 2020-05-18 2020-09-29 中国人民解放军63892部队 PRI (pulse repetition index) conversion radar signal sorting method based on STFT (space time Fourier transform) pre-sorting
CN111693944A (en) * 2020-06-18 2020-09-22 上海志良电子科技有限公司 Radar active interference signal parameter extraction and interference pattern identification method and device
CN112526455A (en) * 2020-12-25 2021-03-19 扬州健行电子科技有限公司 Method for splitting radar-spread-resistant signal into fixed and jittered signals

Non-Patent Citations (3)

* Cited by examiner, † Cited by third party
Title
吴连慧: "基于加权ＳＶＣ和Ｋ－Ｍｅｄｉｏｄｓ联合聚类的", 《舰船电子对抗》 *
吴连慧: "基于改进ＯＰＴＩＣＳ聚类的雷达信号预分选方法", 《舰船电子对抗》 *
张国梁: "雷达信号快速分选和聚类算法研究", 《中国优秀博硕士学位论文全文数据库(硕士)信息科技辑》 *

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116821097A (en) * 2023-07-05 2023-09-29 中国人民解放军92728部队 Full pulse data processing method
CN116821097B (en) * 2023-07-05 2024-01-26 中国人民解放军92728部队 Full pulse data processing method

Also Published As

Publication number Publication date
CN113075637B (en) 2022-06-07

Similar Documents

Publication Publication Date Title
CN110309302B (en) Unbalanced text classification method and system combining SVM and semi-supervised clustering
CN110334591B (en) Unmanned aerial vehicle frequency hopping signal detection and identification method based on cluster analysis
CN110751121B (en) Unsupervised radar signal sorting method based on clustering and SOFM
CN108694346B (en) Ship radiation noise signal identification method based on two-stage CNN
WO2009012056A2 (en) Multiple-instance pruning for learning efficient cascade detectors
CN113705446B (en) Open set identification method for individual radiation source
CN109934203A (en) A kind of cost-sensitive increment type face identification method based on comentropy selection
CN109784270A (en) A kind of processing method promoting face picture identification integrality
CN113075637B (en) Airborne PD radar signal sorting method based on pulse descriptor data compression
CN112528774A (en) Intelligent sorting system and method for unknown radar signals in complex electromagnetic environment
CN110907908A (en) Navigation radar signal sorting method based on envelope analysis
Wiharto et al. Cells identification of acute myeloid leukemia AML M0 and AML M1 using K-nearest neighbour based on morphological images
CN112213697B (en) Feature fusion method for radar deception jamming recognition based on Bayesian decision theory
CN114821335B (en) Unknown target discrimination method based on fusion of depth features and linear discrimination features
CN116911377A (en) Radiation source individual identification method, equipment and medium based on transfer learning
CN116383719A (en) MGF radio frequency fingerprint identification method for LFM radar
CN115809426A (en) Radiation source individual identification method and system
CN112836104B (en) Database-assisted autonomous clustering signal sorting method and system
CN113836526B (en) Intrusion detection method based on improved immune network algorithm and application thereof
CN110826599B (en) Sparse representation sample distribution boundary retention feature extraction method
CN114764879A (en) Aerial target identification method based on radar infrared fusion characteristics
CN110880012B (en) Inter-pulse agile radar radiation source frequency information correlation method for multi-reconnaissance platform
CN109379282B (en) Network community detection method based on multi-label propagation
Zhu et al. Radar signal pre-sorting based on mean-shift algorithm
Liu et al. The minimum description length criterion applied to emitter number detection and pulse classification

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant