CN114839602A - Radar signal statistical feature extraction method and device based on electromagnetic data clustering - Google Patents
Radar signal statistical feature extraction method and device based on electromagnetic data clustering Download PDFInfo
- Publication number
- CN114839602A CN114839602A CN202210776678.3A CN202210776678A CN114839602A CN 114839602 A CN114839602 A CN 114839602A CN 202210776678 A CN202210776678 A CN 202210776678A CN 114839602 A CN114839602 A CN 114839602A
- Authority
- CN
- China
- Prior art keywords
- carrier frequency
- pulse width
- pair
- pulse
- frequency
- 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
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
- G01S7/28—Details of pulse systems
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
- G06F18/232—Non-hierarchical techniques
- G06F18/2321—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/08—Feature extraction
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/12—Classification; Matching
-
- 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)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- Artificial Intelligence (AREA)
- Probability & Statistics with Applications (AREA)
- Computer Networks & Wireless Communication (AREA)
- Radar, Positioning & Navigation (AREA)
- Remote Sensing (AREA)
- Radar Systems Or Details Thereof (AREA)
Abstract
The invention relates to a radar signal statistical characteristic extraction method and device based on electromagnetic data clustering. The method comprises the following steps: calculating the inter-pulse time difference of adjacent pulses according to the arrival time to obtain a time difference-carrier frequency pair set of the pulses; determining a parameter pair with strong aggregation in the time difference-carrier frequency pair set through clustering analysis to obtain a carrier frequency-repeated frequency pair typical value set; checking each parameter pair, determining false repetition frequency, and removing the false repetition frequency from the typical value set of carrier frequency-repetition frequency pairs to obtain a real carrier frequency-repetition frequency pair set; and extracting corresponding pulse width from the radar pulse train based on the real carrier frequency-repetition frequency pair set to obtain a three-parameter feature set of the radar pulse train. The method can greatly weaken the influence of various measurement errors and data noise, and is beneficial to improving the extraction precision of the characteristic parameters; the method realizes the joint extraction of a plurality of characteristic parameters of the radar signal such as carrier frequency, repetition frequency, pulse width and the like, and can more accurately reflect the essential rule of the radar signal.
Description
Technical Field
The invention relates to the technical field of electromagnetic information processing, in particular to a radar signal statistical characteristic extraction method and device based on electromagnetic data clustering.
Background
The radar radiation source is widely applied to solving the problems of target and environment detection and the like, and has a great amount of application in military and civil fields. Radars are usually carried on various land-based, aerial and sea surface weapon platforms and used for detecting information such as surrounding target situation, meteorological hydrological environment and the like and realizing early threat warning and meteorological forecast. The radar works by radiating electromagnetic waves to the space, and the electromagnetic radiation of the radar is exposed to an electronic reconnaissance system while the functions of target detection, environmental perception and the like are realized, so that the radar becomes an important clue for a non-cooperative party to master the attributes of the platform, such as the position, threat, intention and the like of the non-cooperative party. Different radars often adopt different signal parameters according to respective functional requirements, and typical signal parameters include carrier frequency, pulse width, repetition frequency and the like. By utilizing the characteristic, the electronic reconnaissance system can realize the attribute identification of the radar radiation source by intercepting the radar signal and analyzing the statistical characteristics of the radar signal. However, signal parameters of modern radar radiation sources often have a plurality of selectable values, and in an actual electromagnetic environment, significant parameter measurement errors, data noises such as missing pulses and interference pulses and the like usually exist in electronic reconnaissance signals, which causes great difficulty in an extraction process of radar signal statistical characteristics.
The electronic reconnaissance system can obtain a large amount of electromagnetic radiation data of a specific type of radar through accumulation for a period of time, and the data cover various parameter values commonly used by the radar. The electronic reconnaissance data are subjected to cluster analysis, statistical parameters of radar signals are extracted, the working characteristics of various radars and parameter differences among different radars can be reflected visually, and powerful support can be provided for radar model identification. However, the low interception probability of the electronic reconnaissance signal and various measurement errors, data noise and other non-ideal factors significantly increase the difficulty in extracting the statistical parameters of the radar signal.
Because the radar signal interception and analysis problem only exists in a specific application field, the academic papers published at present for the characteristic analysis and extraction problem are few, and because the radar reconnaissance data is difficult to obtain, the related academic research can only be completed based on simulation data and can not truly and comprehensively reflect the situation of actual electronic reconnaissance data, and the proposed radar signal statistical characteristic extraction idea is difficult to avoid to have deviation.
Disclosure of Invention
Therefore, it is necessary to provide a method and an apparatus for extracting statistical characteristics of radar signals based on electromagnetic data clustering, which can eliminate the negative effects of various parameter errors, data noise and other non-ideal factors, accurately estimate typical parameters of radar signal carrier frequency, repetition frequency, pulse width and the like, and achieve the purpose of extracting statistical characteristics of radar signals based on electromagnetic data clustering.
A method for radar signal statistical feature extraction based on electromagnetic data clustering, the method comprising:
obtaining a radar pulse train, the radar pulse train comprising a plurality of radar pulse sub-trains, the radar pulse sub-trains comprising a plurality of pulses, the pulses comprising parameters: carrier frequency, pulse width, arrival time; calculating the inter-pulse time difference of adjacent pulses according to the arrival time to obtain a time difference-carrier frequency pair set of the pulses; determining the parameter pairs with strong aggregation in the time difference-carrier frequency pair set through clustering analysis to obtain a carrier frequency-repeated frequency pair typical value set; checking each parameter pair, determining false repetition frequency, and removing the false repetition frequency from the carrier frequency-repetition frequency pair typical value set to obtain a real carrier frequency-repetition frequency pair set; and extracting corresponding pulse width from the radar pulse train based on the real carrier frequency-repetition frequency pair set to obtain a complete three-parameter feature set of the radar pulse train.
In one embodiment, the method further comprises the following steps: calculating the inter-pulse time difference of adjacent pulses according to the arrival time to obtain a time difference-carrier frequency pair set of the pulses, wherein the method comprises the following steps:
sequentially calculating the time difference between adjacent pulses, screening out a time difference set between the pulses within a repetition frequency value range, and screening out an effective time difference set between the pulses of which tail pulses are within a carrier frequency value range from the time difference set between the pulses; extracting carrier frequency parameters from corresponding pulses based on the effective inter-pulse time difference set, and forming a time difference-carrier frequency pair set with the inter-pulse time difference set; and summarizing the time difference-carrier frequency pairs to form a parameter pair set.
In one embodiment, the method further comprises the following steps: the cluster analysis adopts DBSCAN clustering.
In one embodiment, the method further comprises the following steps: determining the parameter pairs with strong aggregation in the time difference-carrier frequency pair set through clustering analysis to obtain a carrier frequency-repetition frequency pair typical value set, wherein the method comprises the following steps of:
clustering the points with strong centralized aggregation of the parameters through DBSCAN clustering to obtain effective clustering serial numbers of the parameters; and selecting corresponding parameter pairs from the time difference-carrier frequency pair set according to the effective clustering serial numbers of the parameter pairs to respectively form parameter pair subsets, and averaging the parameters in the parameter pair subsets to obtain a carrier frequency-repeated frequency pair typical value set.
In one embodiment, the method further comprises the following steps: checking each parameter pair, determining false repetition frequency, and removing the false repetition frequency from the carrier frequency-repetition frequency pair typical value set to obtain a real carrier frequency-repetition frequency pair set, which comprises the following steps:
arranging the repetition frequency parameters in the typical value set of carrier frequency-repetition frequency pairs from small to large to form a repetition frequency vector, and determining the maximum possible continuous pulse missing number; the maximum possible continuous missing pulse number is obtained by carrying out quotient rounding-up calculation according to the upper limit value of the repetition frequency value range and the maximum value of the repetition frequency parameter in the repetition frequency vector; checking each parameter pair according to the error allowable range, and determining the parameter pair as a false repetition frequency if the difference value between the value of the parameter pair and a certain element in the time difference set is within the measurement error allowable range; the elements in the time difference set are obtained by carrying out integral multiple weighted summation according to any two repetition frequency parameters in the repetition frequency vector; and eliminating the carrier frequency-repeated frequency pair typical value corresponding to the false repeated frequency from the carrier frequency-repeated frequency pair typical value set to obtain a real carrier frequency-repeated frequency pair set.
In one embodiment, the method further comprises the following steps: extracting a corresponding pulse width, comprising:
judging whether the radar works in a specific carrier frequency and a repetition frequency mode, if so, extracting corresponding pulse width parameters according to a simple averaging method; if not, the significant outliers are removed from the pulse width set, and the pulse width measurement values with the significant outliers removed are subjected to cluster analysis.
In one embodiment, the method further comprises the following steps: removing the remarkable outliers from the pulse width measurement values, and performing cluster analysis on the pulse width measurement values with the remarkable outliers removed, wherein the cluster analysis comprises the following steps:
extracting pulse widths corresponding to the real carrier frequency-repetition frequency pair set from the pulses to form a pulse width set; eliminating parameters smaller than the minimum possible value of the pulse width in the pulse width set to obtain a pulse width measured value set; the minimum possible value of the pulse width is set according to the prior knowledge of the radar signal parameters under the condition of considering the measurement error; summarizing the pulse width measurement values to form a pulse width point set; clustering the points with strong aggregation in the pulse width point set by DBSCAN clustering to obtain effective pulse width clusters consisting of the points with strong aggregation and serial numbers thereof; and extracting corresponding pulse width parameters from the pulse width measurement value set according to the pulse width effective clustering sequence number to form a pulse width subset, and averaging the parameters in the pulse width subset to obtain a typical pulse width value set.
In one embodiment, the method further comprises the following steps: extracting corresponding pulse width from a radar pulse train based on the real carrier frequency-repetition frequency pair set to obtain a three-parameter characteristic set of the pulse, wherein the three-parameter characteristic set comprises the following steps:
combining the pulse width typical value set with the real carrier frequency-repetition frequency pair set to obtain a three-parameter typical set; and traversing the carrier frequency-repetition frequency pair typical value set, and respectively counting the pulse width parameters of the radar pulse trains to obtain a complete three-parameter feature set of the radar pulse trains.
In one embodiment, the method further comprises the following steps: clustering by using DBSCAN, comprising: when DBSCAN clustering analysis is carried out on the point set by the parameters, the clustering radius is set to be 1, and the clustering frequency threshold is set according to the pulse number of the electronic reconnaissance radar; when the DBSCAN clustering analysis is carried out on the pulse pair point set, the clustering radius is set to be 3 times of the pulse width measurement standard deviation, and the clustering frequency threshold is set according to the pulse number of the electronic reconnaissance radar.
A radar signal statistical feature extraction apparatus based on electromagnetic data clustering, the apparatus comprising:
a data acquisition module configured to acquire a radar pulse train, the radar pulse train including a plurality of radar pulse sub-trains, the radar pulse sub-trains including a plurality of pulses, the pulses including parameters: carrier frequency, pulse width, arrival time.
The cluster analysis module is used for calculating the inter-pulse time difference of adjacent pulses according to the arrival time to obtain a time difference-carrier frequency pair set of the pulses; and determining the parameter pairs with stronger aggregation in the time difference-carrier frequency pair set through clustering analysis to obtain a carrier frequency-repeated frequency pair typical value set.
And the checking module is used for checking each parameter pair, determining false repetition frequency, and removing the false repetition frequency from the carrier frequency-repetition frequency pair typical value set to obtain a real carrier frequency-repetition frequency pair set.
And the pulse width parameter extraction module is used for extracting corresponding pulse widths from the radar pulse train based on the real carrier frequency-repetition frequency pair set to obtain a three-parameter feature set of the pulses.
According to the radar signal statistical characteristic extraction method and device based on electromagnetic data clustering, the parameter pair with strong aggregation in the time difference-carrier frequency pair set is determined through clustering analysis, and the carrier frequency-repetition frequency pair typical value set is obtained, so that the influence of various measurement errors and data noise can be greatly weakened, and wild values which are obviously deviated from real parameters are excluded from the statistical analysis process, so that the adaptability to the mixed electromagnetic environment is enhanced, and the characteristic parameter extraction precision is improved. Each parameter pair is checked to determine the false repetition frequency, the false repetition frequency is removed from the carrier frequency-repetition frequency pair typical value set to obtain a real carrier frequency-repetition frequency pair set, and the phenomenon that a high-order repetition frequency value caused by pulse leakage is mistaken for a real repetition frequency parameter is avoided. And finally, extracting corresponding pulse width from the radar pulse train based on the real carrier frequency-repetition frequency pair set to obtain a three-parameter feature set of the pulse, so that the joint extraction of multiple feature parameters such as the carrier frequency, the repetition frequency and the pulse width of the radar signal is realized, instead of extracting each one-dimensional feature after artificially segmenting the mutual relation among the multiple parameters, the multi-parameter joint feature can more accurately reflect the essential rule of the radar signal, the scale of the feature set of the radar signal can be effectively compressed, and the method is favorable for supporting various applications based on the statistical parameter feature.
Drawings
FIG. 1 is a flowchart of a radar signal statistical feature extraction method based on electromagnetic data clustering according to the present invention;
FIG. 2 is a scatter diagram of observation samples of a carrier frequency-arrival time two-dimensional parameter of an original radar signal in one embodiment;
FIG. 3 is a two-dimensional dispersion chart of carrier frequency-time difference of the radar signal screened according to the carrier frequency and repetition frequency range in FIG. 2;
FIG. 4 is a radar signal pulse width distribution histogram corresponding to a single sample parameter obtained after cluster analysis is performed on the sample signals in FIG. 3;
FIG. 5 is a scatter diagram of three-dimensional parameters of carrier frequency, repetition frequency and pulse width of the radar signal in FIG. 4; wherein, (a) is a three-dimensional space scatter diagram of carrier frequency-repetition frequency-pulse width; (b) the method is a scatter diagram of radar signal carrier frequency-repetition frequency-pulse width three-dimensional parameters.
Detailed Description
In order to make the objects, technical solutions and advantages of the present application more apparent, the present application is described in further detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the present application and are not intended to limit the present application.
The invention provides a radar signal statistical characteristic extraction method and device based on electromagnetic data clustering, which take a certain amount of electronic reconnaissance data as a basis, decompose a radar signal statistical characteristic extraction process into a parameter pair (consisting of carrier frequency and repetition frequency) and a pulse width extraction process, extract a parameter pair with strong clustering property by means of data clustering, and eliminate the influence of stray data noise and random measurement errors; then, checking the clustering result to avoid mistakenly considering the false repetition frequency as a real repetition frequency parameter; and finally, extracting corresponding pulse width to obtain a radar signal three-parameter feature set.
It should be noted that the "-" appearing in the time difference-carrier frequency, carrier frequency-repetition frequency-pulse width of the present invention indicates a parameter association symbol.
As shown in fig. 1, a flowchart of a method for extracting statistical features of radar signals based on electromagnetic data clustering provided by the present invention includes the following steps:
s1: acquiring a radar pulse train, wherein the radar pulse train comprises a plurality of radar pulse subsequences, the radar pulse subsequences comprise a plurality of pulses, and the pulses comprise parameters: carrier frequency, pulse width, arrival time.
Specifically, the number of radar pulse subcolumns in the radar pulse train intercepted by electronic reconnaissance is set as
The signal arrival time sequence is
Corresponding signal carrier frequencies are respectively
。
S2: calculating the inter-pulse time difference of adjacent pulses according to the arrival time to obtain a time difference-carrier frequency pair set of the pulses; and determining the parameter pairs with stronger aggregation in the time difference-carrier frequency pair set through clustering analysis to obtain a carrier frequency-repeated frequency pair typical value set.
Specifically, the time difference between adjacent pulses is sequentially calculated, a time difference set between the pulses within a repetition frequency value range is screened out, and an effective time difference set between the pulses with tail pulses within a carrier frequency value range is screened out from the time difference set between the pulses; extracting carrier frequency parameters from corresponding pulses based on the effective inter-pulse time difference set to form a carrier frequency measurement value set, and forming a time difference-carrier frequency pair set with the pulse time difference set; and summarizing the time difference-carrier frequency pairs to form a parameter pair set.
Clustering the points with stronger aggregation in the parameter pair point set through clustering analysis to obtain effective clustering sequence numbers of the parameter pairs; and selecting corresponding parameter pairs from the time difference-carrier frequency pair set according to the effective clustering serial numbers of the parameter pairs to respectively form parameter pair subsets, and averaging the parameters in the parameter pair subsets to obtain a carrier frequency-repeated frequency pair typical value set.
S3: and checking each parameter pair, determining false repetition frequency, and removing the false repetition frequency from the carrier frequency-repetition frequency pair typical value set to obtain a real carrier frequency-repetition frequency pair set.
Specifically, the repetition frequency parameters in the carrier frequency-repetition frequency pair typical value set are arranged from small to large to form a repetition frequency vector, and the maximum possible continuous pulse missing number is determined; the maximum possible number of continuous missing pulses is obtained by carrying out quotient rounding-up calculation according to the upper limit value of the repetition frequency value range and the maximum value of the repetition frequency parameter in the repetition frequency vector; checking each parameter pair according to the error allowable range, and determining the parameter pair as a false repetition frequency if the difference value between the value of the parameter pair and a certain element in the time difference set is within the measurement error allowable range; elements in the time difference set are obtained by carrying out weighted summation according to any two repetition frequency parameters in the repetition frequency vector; and removing the parameter pair typical value corresponding to the false repetition frequency from the carrier frequency-repetition frequency pair typical value set to obtain a real carrier frequency-repetition frequency pair set.
S4: and extracting corresponding pulse width from the radar pulse train based on the real carrier frequency-repetition frequency pair set to obtain a three-parameter feature set of the radar pulse train.
Specifically, whether the radar works in a specific carrier frequency and repetition frequency mode is judged, and if yes, corresponding pulse width parameters are extracted according to a simple averaging method; if not, the significant outliers are removed from the pulse width set, and the pulse width measurement values with the significant outliers removed are subjected to cluster analysis.
Extracting pulse widths corresponding to the real carrier frequency-repetition frequency pair set from the pulses to form a pulse width set; eliminating the parameters smaller than the minimum possible value of the pulse width in the pulse width set to obtain a pulse width measured value set; the minimum possible value of the pulse width is set according to the prior knowledge of the radar signal parameters under the condition of considering the measurement error; summarizing the pulse width measurement values to form a pulse width point set; clustering the points with strong pulse width point concentration through clustering to obtain effective pulse width clusters and serial numbers thereof consisting of the points with strong aggregation; and extracting corresponding pulse width parameters from the pulse width measurement value set according to the pulse width effective clustering sequence number to form a pulse width subset, and averaging the parameters in the pulse width subset to obtain a typical pulse width value set.
Combining the pulse width typical value set with a real carrier frequency-repetition frequency pair set to obtain a three-parameter typical value set; and traversing the carrier frequency-repetition frequency pair typical value set, and respectively counting the pulse width parameters of the radar pulse sub-train corresponding to each carrier frequency-repetition frequency parameter pair to obtain a three-parameter characteristic set of the radar pulse train.
According to the radar signal statistical characteristic extraction method and device based on electromagnetic data clustering, the parameter pair with strong aggregation in the time difference-carrier frequency pair set is determined through clustering analysis, and the carrier frequency-repetition frequency pair typical value set is obtained, so that the influence of various measurement errors and data noise can be greatly weakened, and wild values which are obviously deviated from real parameters are excluded from the statistical analysis process, so that the adaptability to the mixed electromagnetic environment is enhanced, and the characteristic parameter extraction precision is improved. Each parameter pair is checked to determine the false repetition frequency, the false repetition frequency is removed from the carrier frequency-repetition frequency pair typical value set to obtain a real carrier frequency-repetition frequency pair set, and the phenomenon that a high-order repetition frequency value caused by pulse leakage is mistaken for a real repetition frequency parameter is avoided. And finally, extracting corresponding pulse width from the radar pulse train based on the real carrier frequency-repetition frequency pair set to obtain a three-parameter feature set of the pulse, so that joint extraction of multiple feature parameters of the radar signal carrier frequency, repetition frequency, pulse width and the like is realized, instead of extracting each one-dimensional feature after artificially splitting the mutual relation among the multiple parameters, the multi-parameter joint feature can more accurately reflect the essential rule of the radar signal, can effectively compress the scale of the feature set of the radar signal, and is favorable for supporting various applications based on statistical parameter features.
In one embodiment, calculating the inter-pulse time difference of adjacent pulses according to the arrival time to obtain a set of time difference-carrier frequency pairs of the pulses, comprises:
sequentially calculating the time difference between adjacent pulses, screening out a time difference set between the pulses within a repetition frequency value range, and screening out an effective time difference set between the pulses with tail pulses within a carrier frequency value range from the time difference set between the pulses; extracting carrier frequency parameters from corresponding pulses based on the effective inter-pulse time difference set to form a carrier frequency measurement value set, and forming a time difference-carrier frequency pair set with the pulse time difference set; and summarizing the time difference-carrier frequency pairs to form a parameter pair set.
Specifically, the repetition frequency value range of the radar signal is set as follows according to experience
Setting the carrier frequency range of the radar signal as
。
Sequentially calculating the inter-pulse time difference of the pulses with closer arrival time
Screening out the time difference between pulses
Internal and tail pulse carrier frequency is located
The effective time differences in the interior form a set of effective inter-pulse time differences
The carrier frequency measurements of the tail pulses in the respective pulse pair form a set
，
And
the corresponding element in (1) constitutes a parameter pair
Set, in time difference-carrier frequency (DTOA-RF) two-dimensionalForming a point set on the plane;
in one embodiment, a cluster analysis method is further provided, which specifically includes: and clustering by adopting a DBSCAN function.
It is worth to be noted that the DBSCAN algorithm determines the clustering strategy of each sample by calculating the distance between each sample and the clustering center, but the invention measures the similarity of the parameters of the scattered pulses of the same radar in the time domain to extract the common characteristics of the samples, and the DBSCAN algorithm is selected in consideration of the fact that the radar pulse parameters intercepted by electronic reconnaissance are the result of the real radar signal parameters added with random observation noise and the observation errors of different pulse parameters are independent of each other.
In one embodiment, determining a highly aggregated parameter pair in the time difference-carrier frequency pair set by cluster analysis to obtain a carrier frequency-repetition frequency pair typical value set includes:
clustering the points with strong centralized aggregation of the parameters through DBSCAN clustering to obtain effective clustering serial numbers of the parameters; and selecting corresponding parameter pairs from the time difference-carrier frequency pair set according to the effective clustering serial numbers of the parameter pairs to respectively form parameter pair subsets, and averaging the parameters in the parameter pair subsets to obtain a carrier frequency-repeated frequency pair typical value set.
Specifically, clustering is carried out through a DBSCAN function, the clustering radius is set to be 1, and the threshold value of the clustering frequency is set to be 1 according to the pulse number of the electronic reconnaissance radar
Realize to the parameter point-to-point set
Strong medium aggregation (number of observation samples is larger than that of observation samples)
) The clustering of the point set outputs the clustering serial number corresponding to each sample, namely the effective clustering serial number, and outputs field value identification-1 for the non-clustered discrete points; according to valid cluster sequence number
From parameter to point set
Corresponding time difference-carrier frequency samples screened in the process are respectively formed into a set
Averaging the sample parameters in each subset to a corresponding set of carrier frequency-repetition frequency pair representative values
。
In one embodiment, the checking each parameter pair, determining a spurious repetition frequency, and removing the spurious repetition frequency from the set of carrier frequency-repetition frequency pair typical values to obtain a set of real carrier frequency-repetition frequency pairs, includes:
arranging the repetition frequency parameters in the typical value set of carrier frequency-repetition frequency pairs from small to large to form a repetition frequency vector, and determining the maximum possible continuous pulse missing number; the maximum possible continuous missing pulse number is obtained by carrying out quotient rounding calculation according to the upper limit value of the repetition frequency value range and the maximum value of the repetition frequency parameter in the repetition frequency vector; checking each parameter pair according to the error allowable range, and determining the parameter pair as a false repetition frequency if the difference value between the value of the parameter pair and a certain element in the time difference set is within the measurement error allowable range; removing the carrier frequency-repeated frequency pair typical value corresponding to the false repeated frequency from the carrier frequency-repeated frequency pair typical value set to obtain a real carrier frequency-repeated frequency pair set; and the elements in the time difference set are obtained by carrying out integral multiple weighted summation according to any two repetition frequency parameters in the repetition frequency vector.
In particular, the repetition frequency interval value
The vectors are arranged from small to large to form a vector
Determining the maximum possible number of consecutive missing pulses
Wherein, in the step (A),
representing a vector
The maximum value of the medium repetition frequency parameter,
indicating rounding up.
Will vector
Weighting and summing any two elements to obtain a time difference set
Relative amount of
Each parameter in (1) is checked one by one, and if the value of each parameter is equal to the time difference set
If the difference value of a certain element is within the allowable error range, for example, less than 3 times of standard deviation of time of arrival measurement, said time difference value is considered as false time difference caused by missing pulse, so that the parameter corresponding to said time difference value is used as typical value
And removing from the set of representative values of carrier frequency-repetition frequency pairs.
For example, for a radar pulse train with a true repetition frequency of 200us, when the pulse missing probability is high, a large number of pulse arrival time differences of 400us appear in the pulse train, the repetition frequency parameter set obtained by parameter statistics is 200us, 400us, and the step of "removing false repetition frequency" can remove the second-order repetition frequency value of 400us by using the multiple relation of the values of the two parameters, and finally only outputs the true parameter of 200 us.
In one embodiment, a method for extracting a corresponding pulse width is provided, which specifically includes: judging whether the radar works in a specific carrier frequency and repetition frequency mode, if so, extracting corresponding pulse width parameters according to a simple averaging method; if not, the significant outliers are removed from the pulse width set, and the pulse width measurement values with the significant outliers removed are subjected to cluster analysis.
It should be noted that when the radar operates in a specific carrier frequency and repetition frequency mode, the pulse width value is a determined value, so that the pulse width can be statistically analyzed by a simple averaging method, and the pulse width extraction process can be simplified. When the pulse width mode of the radar signal is complex, or the pulse width measurement process has problems of significant pulse splitting and the like, significant outliers need to be removed from the pulse width set, and the cluster analysis is performed on the pulse width measurement values with the significant outliers removed; the accuracy of the pulse width extraction can be improved.
In one embodiment, the removing significant outliers from the pulse width measurements, and performing cluster analysis on the pulse width measurements from which significant outliers are removed includes:
extracting pulse widths corresponding to the real carrier frequency-repetition frequency pair set from the pulses to form a pulse width set; eliminating the parameters smaller than the minimum possible value of the pulse width in the pulse width set to obtain a pulse width measured value set; the minimum possible value of the pulse width is set according to the prior knowledge of the radar signal parameters under the condition of considering the measurement error; summarizing the pulse width measurement values to form a pulse width point set; clustering points with strong pulse width point concentration through DBSCAN clustering to obtain effective pulse width clusters and serial numbers thereof formed by the points; and extracting corresponding pulse width parameters from the pulse width measurement value set according to the pulse width effective clustering sequence number to form a pulse width subset, and averaging the parameters in the pulse width subset to obtain a typical pulse width value set.
In particular, pairs of extracted and true parameters
Corresponding radar signal pulse width values form a pulse width set
Setting the minimum possible value of the signal pulse width (considering the measurement error) according to the prior knowledge of the radar signal parameter
Integrating the pulse width
Median value of less than
The measured values are removed as erroneous observations caused by pulse splitting and the like, and a set of pulse width measured values is obtained and recorded as
. The minimum value of the pulse width is equal to the minimum nominal value of the pulse width of the radar signal minus 3 times of the measurement standard deviation, i.e. the minimum value of the pulse width = minimum nominal value of the pulse width of the radar signal-measurement standard deviation x 3.
Clustering the radar signal pulse width measurement set through the DBSCAN function, setting the clustering radius to be 3 times of the pulse width measurement standard deviation, and setting the clustering frequency threshold value to be 3 times of the pulse number of the electronic reconnaissance radar
Implementing the data set
Strong medium aggregation (number of observation samples is larger than that of observation samples)
) And (3) clustering the point set, outputting the clustering serial number corresponding to each sample, and outputting field value identification-1 for the non-clustered discrete points.
According to valid cluster sequence number
From a data set
Screening corresponding pulse width samples to respectively form pulse width subsets
Averaging sample parameters in each pulse width subset to obtain a corresponding pulse width typical value set
。
In one embodiment, extracting corresponding pulse widths from a radar pulse train based on a set of real carrier frequency-repetition frequency pairs to obtain a three-parameter feature set of pulses includes:
combining the pulse width typical value set with a real carrier frequency-repetition frequency pair set to obtain a three-parameter typical value set; and traversing the carrier frequency-repetition frequency pair typical value set, and respectively counting the pulse width parameters of the radar pulse trains to obtain a complete three-parameter feature set of the radar pulse trains.
Specifically, the statistical result of the radar signal pulse width parameter is integrated with the corresponding set of real carrier frequency-repetition frequency pairs to obtain a three-parameter typical set of the radar signal corresponding to the parameter, wherein the three-parameter typical set is composed of the carrier frequency, the repetition frequency and the pulse width
。
Traversing the obtained carrier frequency-repeat frequency pair typical value set
And respectively counting pulse width parameters of corresponding radar signals to obtain a complete three-parameter feature set.
It should be reminded that, in this embodiment, the method for extracting a typical pulse width value is described with respect to a specific carrier frequency-repetition frequency value, and since a radar signal may have a plurality of typical carrier frequency-repetition frequency parameters, the typical pulse width value is applied to each carrier frequency-repetition frequency parameter ("traversal"), and finally a complete set of three parameter typical values of carrier frequency-repetition frequency-pulse width is obtained.
It should be noted that the parameters commonly used for radar pulse include carrier frequency, pulse width, repetition frequency, etc., where the carrier frequency and repetition frequency have higher reliability, the pulse width is affected by various factors to have larger measurement error, and the measurement error caused by pulse splitting is relatively low in reliability. In order to solve the problem of extraction of radar signal parameter characteristics, although input data contains a plurality of pieces of radar data, the situation that the input data are staggered in a time domain is less, and the time difference of adjacent pulses is directly calculated to obtain obvious repeated frequency characteristics, so that carrier frequencies and repeated frequencies are selected as clustering parameters, and finally, the pulse width characteristics of signals are counted in each clustering result. And finally, obtaining a complete three-parameter feature set.
In one embodiment, the DBSCAN function is mainly used for clustering, and the specific content is as follows:
and when the DBSCAN clustering analysis is carried out on the point set by the parameters, the clustering radius is set to be 1, and the clustering frequency threshold is set according to the pulse number of the electronic reconnaissance radar.
When DBSCAN clustering analysis is carried out on the pulse pair point set, the clustering radius is set to be 3 times of the pulse width measurement standard deviation, and the clustering frequency threshold is set according to the pulse number of the electronic reconnaissance radar.
It should be noted that the threshold value of the clustering frequency is related to the number of radar pulses, the magnitude of data noise, the complexity of radar repetition frequency mode, and the like. Empirically, the threshold of desirable clustering frequency is equal to between 1/100 and 1/20 of the number of radar pulses.
The standard deviation of the pulse width measurement depends on factors such as the level of equipment and the strength of signals. Empirically, the standard deviation can be taken to be 0.05 microseconds to 0.1 microseconds.
It should be understood that, although the steps in the flowchart of fig. 1 are shown in order as indicated by the arrows, the steps are not necessarily performed in order as indicated by the arrows. The steps are not performed in the exact order shown and described, and may be performed in other orders, unless explicitly stated otherwise. Moreover, at least a portion of the steps in fig. 1 may include multiple sub-steps or multiple stages that are not necessarily performed at the same time, but may be performed at different times, and the order of performance of the sub-steps or stages is not necessarily sequential, but may be performed in turn or alternately with other steps or at least a portion of the sub-steps or stages of other steps.
In one embodiment, there is provided an electromagnetic data clustering-based radar signal statistical feature extraction apparatus, including:
a data acquisition module for acquiring a radar pulse train, the radar pulse train comprising a plurality of radar pulse trains, the radar pulse trains comprising a plurality of pulses, the pulses comprising parameters: carrier frequency, pulse width, arrival time.
The cluster analysis module is used for calculating the inter-pulse time difference of adjacent pulses according to the arrival time to obtain a time difference-carrier frequency pair set of the pulses; and determining the parameter pairs with stronger aggregation in the time difference-carrier frequency pair set through clustering analysis to obtain a carrier frequency-repeated frequency pair typical value set. And the checking module is used for checking each parameter pair, determining the false repetition frequency, and removing the false repetition frequency from the carrier frequency-repetition frequency pair typical value set to obtain a real carrier frequency-repetition frequency pair set.
And the pulse width parameter extraction module is used for extracting corresponding pulse widths from the radar pulse train based on the real carrier frequency-repetition frequency pair set to obtain a three-parameter feature set of the pulses.
For specific limitations of the electromagnetic data clustering-based radar signal statistical feature extraction device, reference may be made to the above limitations of the electromagnetic data clustering-based radar signal statistical feature extraction method, and details thereof are not repeated here. All or part of the modules in the electromagnetic data clustering-based radar signal statistical characteristic extraction device can be realized by software, hardware and a combination thereof. The modules can be embedded in a hardware form or independent from a processor in the computer device, and can also be stored in a memory in the computer device in a software form, so that the processor can call and execute operations corresponding to the modules.
Fig. 2 to 5 show exemplary processing results of radar reconnaissance data obtained from a certain actual measurement experiment by using the method provided by the present invention, and experimental parameters are set as follows:
the radar to be analyzed works in the C frequency band, and the effective carrier frequency range is
The effective repetition frequency interval range is
The electronic reconnaissance signal of the radar is mixed with the other X-band radar and one Ka-band radar.
Fig. 2 shows the arrival time of the radar signal captured by the electronic reconnaissance receiver in about 0.2 second and the distribution of the carrier frequency parameters, and it can clearly see the distribution rule of the signal of the C, X, Ka frequency band 3 radars in the frequency domain, so that the carrier frequency parameters can be used to separate the signals of the 3 radars. However, for the C-band radar signal to be analyzed, due to the introduction of significant parameter measurement errors and a large amount of data noise in the electronic reconnaissance process, the laws of carrier frequency, repetition frequency and the like are not visually presented in the observation data, and are difficult to directly obtain by adopting a simple data averaging mode. The complexity of the data case in this diagram illustrates the necessity of analyzing the electronic reconnaissance data in depth to obtain statistical characteristics of the radar signal.
FIG. 3 shows the electron scout pulse train of FIG. 2 at a carrier frequency range
And repetition interval range
After internal screening, a carrier frequency-time difference two-dimensional scatter diagram of the pulse train is reserved. It can be seen that the radar signal exhibits a certain regularity within the parameter range, and the inter-pulse time difference and the signal carrier frequency are mainly concentrated on a series of discrete parameter values. However, it is affected by measurement errors and data noise introduced during signal acquisition, and by the large number of (non-adjacent) multiple-stage pulse interval calculations introduced during the inter-pulse time difference calculationAs a result, the observation sample derives many spurious features on the basis of the true signal features. As shown in fig. 3, the outliers with weak data aggregation and spurious features corresponding to repetition frequency splitting and high-order repetition frequency caused by missing pulses, interference pulses, and other factors are shown. The influence of measurement errors and data noise is eliminated by an effective method, and then the real radar signal statistical characteristics can be obtained.
Fig. 4 is a radar signal pulse width distribution histogram corresponding to a single sample parameter obtained after cluster analysis is performed on the sample signals in fig. 3. During clustering, under the conditions that the radius of a clustering ellipse is set to be 3 and the threshold of clustering frequency is set to be 1000, clustering analysis is carried out on the signal samples in the graph 3 to obtain 21 classes of effective samples, and outlier points which do not form effective clusters are marked to be '1' and removed. It can be seen that, with the help of cluster analysis, interference of various measurement errors and data noise is eliminated, and after partial data with strong aggregation is extracted from stray observation data, the observation data sets are effectively distributed in a small neighborhood centered on the real parameters of radar signals and are approximately normally distributed under the influence of random noise.
The observation data of the class 1 signal sample is screened from the original reconnaissance data, the pulse width parameters of the corresponding radar signals are extracted, and histogram statistics is performed to obtain the result shown in fig. 4. The pulse width of the radar signal is mostly distributed between 2.8 microseconds and 3 microseconds in the figure, which shows that when the radar works by using the carrier frequency and the repetition frequency parameter, only 1 typical pulse width value is probably used, and is about 2.86 microseconds. There are also few radar signals in fig. 4 with pulse width measurements close to 0, which is a false observation caused by pulse splitting. Setting the lower limit of the pulse width of the radar signal to be 1 microsecond to eliminate an error observation value introduced by pulse splitting, setting the radius of a DBSCAN clustering function to be 0.2 microsecond, setting the threshold value of the number of clustering samples to be 300, gathering the screened radar pulse width parameters into a class, directly averaging the pulse width of the radar signal in the cluster, and taking the pulse width mean value as the statistical characteristic of the radar pulse width.
Fig. 5 is a distribution of radar signal carrier frequency-repetition frequency-pulse width three-dimensional statistical characteristics obtained after statistical analysis is performed on all 21 carrier frequency-repetition frequency parameters of the clustering results in fig. 4, where (a) is a carrier frequency-repetition frequency-pulse width three-dimensional spatial scatter diagram, and (b) is a projection result on a carrier frequency-repetition frequency two-dimensional plane. (a) The pulse width parameters of the cluster characteristics are very close to each other, are intensively dispersed in the range of 2.82 microseconds to 2.9 microseconds, and only small deviation is introduced by observation errors, which indicates that the radar mainly adopts a single pulse width mode, and the pulse width value is close to 2.9 microseconds. (b) The carrier frequency-repetition frequency two-dimensional plane projection result in (1) shows that the radar mainly adopts 6 typical carrier frequencies and 4 typical repetition frequency values, wherein the 6 carrier frequency values are respectively about 5300MHz, 5360MHz, 5420MHz, 5480MHz, 5540MHz and 5660MHz, and the 4 repetition frequency values are respectively about 384 mu s, 416 mu s, 448 mu s and 480 mu s. The signal parameter features in the observed data are not simple combinations of typical values of these one-dimensional features, where 3 carrier-frequency-repetition parameter combinations (5300 MHz, 384 μ s), (5420 MHz, 480 μ s) and (5660 MHz, 480 μ s) do not form a good data cluster, probably because the radar does not use or uses less of these several operating modes during the reconnaissance period, and also because the radar does not have these several operating modes. The unique signal characteristic rules cannot be obtained by using mutually independent one-dimensional characteristic parameter analysis methods, and can only be obtained by means of the multi-dimensional parameter clustering provided by the invention.
It will be understood by those skilled in the art that all or part of the processes of the methods of the embodiments described above may be implemented by hardware instructions of a computer program, which may be stored in a non-volatile computer-readable storage medium, and when executed, may include the processes of the embodiments of the methods described above. Any reference to memory, storage, database, or other medium used in the embodiments provided herein may include non-volatile and/or volatile memory, among others. Non-volatile memory can include read-only memory (ROM), Programmable ROM (PROM), Electrically Programmable ROM (EPROM), Electrically Erasable Programmable ROM (EEPROM), or flash memory. Volatile memory can include Random Access Memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in a variety of forms, such as Static RAM (SRAM), Dynamic RAM (DRAM), Synchronous DRAM (SDRAM), Double Data Rate SDRAM (DDRSDRAM), Enhanced SDRAM (ESDRAM), synchronous Link (Synchlink) DRAM (SLDRAM), Rambus Direct RAM (RDRAM), direct bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).
The technical features of the above embodiments can be arbitrarily combined, and for the sake of brevity, all possible combinations of the technical features in the above embodiments are not described, but should be considered as the scope of the present specification as long as there is no contradiction between the combinations of the technical features.
The above-mentioned embodiments only express several embodiments of the present application, and the description thereof is more specific and detailed, but not construed as limiting the scope of the invention. It should be noted that, for a person skilled in the art, several variations and modifications can be made without departing from the concept of the present application, which falls within the scope of protection of the present application. Therefore, the protection scope of the present patent shall be subject to the appended claims.
Claims (10)
1. A radar signal statistical feature extraction method based on electromagnetic data clustering is characterized by comprising the following steps:
obtaining a radar pulse train, the radar pulse train comprising a plurality of radar pulse sub-trains, the radar pulse sub-trains comprising a plurality of pulses, the pulses comprising parameters: carrier frequency, pulse width, arrival time;
calculating the inter-pulse time difference of adjacent pulses according to the arrival time to obtain a time difference-carrier frequency pair set of the pulses; determining the parameter pairs with strong aggregation in the time difference-carrier frequency pair set through clustering analysis to obtain a carrier frequency-repeated frequency pair typical value set;
checking each parameter pair, determining false repetition frequency, and removing the false repetition frequency from the carrier frequency-repetition frequency pair typical value set to obtain a real carrier frequency-repetition frequency pair set;
and extracting corresponding pulse width from the radar pulse train based on the real carrier frequency-repetition frequency pair set to obtain a complete three-parameter feature set of the radar pulse train.
2. The method of claim 1, wherein computing inter-pulse time differences of adjacent pulses from the arrival times to obtain a set of time difference-carrier frequency pairs for the pulses comprises:
sequentially calculating the time difference between adjacent pulses, screening out a time difference set between the pulses within a repetition frequency value range, and screening out an effective time difference set between the pulses of which tail pulses are within a carrier frequency value range from the time difference set between the pulses;
extracting carrier frequency parameters from corresponding pulses based on the effective inter-pulse time difference set, and forming a time difference-carrier frequency pair set with the inter-pulse time difference set;
and summarizing the time difference-carrier frequency pairs to form a parameter pair set.
3. The method according to claim 1 or 2, characterized in that:
the cluster analysis adopts DBSCAN clustering.
4. The method of claim 3, wherein determining the more highly aggregated parameter pairs in the set of TD-RCs by cluster analysis to obtain a set of exemplary carrier-RF pairs comprises:
clustering the points with strong centralized aggregation of the parameters through DBSCAN clustering to obtain effective clustering serial numbers of the parameters;
and selecting corresponding parameter pairs from the time difference-carrier frequency pair set according to the effective clustering serial numbers of the parameter pairs to respectively form parameter pair subsets, and averaging the parameters in the parameter pair subsets to obtain a carrier frequency-repeated frequency pair typical value set.
5. The method of claim 1, wherein checking each pair of parameters to determine a spurious repetition frequency, and removing the spurious repetition frequency from the set of carrier frequency-repetition frequency pair representative values to obtain a set of true carrier frequency-repetition frequency pairs comprises:
arranging the repetition frequency parameters in the typical value set of carrier frequency-repetition frequency pairs from small to large to form a repetition frequency vector, and determining the maximum possible continuous pulse missing number; the maximum possible continuous missing pulse number is obtained by carrying out quotient rounding-up calculation according to the upper limit value of the repetition frequency value range and the maximum value of the repetition frequency parameter in the repetition frequency vector;
checking each parameter pair according to the error allowable range, and determining the parameter pair as a false repetition frequency if the difference value between the value of the parameter pair and a certain element in the time difference set is within the measurement error allowable range; the elements in the time difference set are obtained by carrying out integral multiple weighted summation according to any two repetition frequency parameters in the repetition frequency vector;
and eliminating the carrier frequency-repeated frequency pair typical value corresponding to the false repeated frequency from the carrier frequency-repeated frequency pair typical value set to obtain a real carrier frequency-repeated frequency pair set.
6. The method of claim 3, wherein extracting respective pulse widths comprises:
judging whether the radar works in a specific carrier frequency and repetition frequency mode, if so, extracting corresponding pulse width parameters according to a simple averaging method;
if not, the significant outliers are removed from the pulse width set, and the measured pulse width values with the significant outliers removed are subjected to cluster analysis.
7. The method of claim 6, wherein significant outliers are removed from the pulse width measurements, and performing cluster analysis on the pulse width measurements from which significant outliers are removed comprises:
extracting pulse widths corresponding to the real carrier frequency-repetition frequency pair set from the pulses to form a pulse width set; parameters smaller than the minimum possible value of the pulse width in the pulse width set are removed to obtain a pulse width measured value set; the minimum possible value of the pulse width is set according to the prior knowledge of the radar signal parameters under the condition of considering the measurement error;
summarizing the pulse width measurement values to form a pulse width point set;
clustering the points with strong aggregation in the pulse width point set by DBSCAN clustering to obtain effective pulse width clusters consisting of the points with strong aggregation and serial numbers thereof;
and extracting corresponding pulse width parameters from the pulse width measurement value set according to the pulse width effective clustering sequence numbers to form a pulse width subset, and averaging the parameters in the pulse width subset to obtain a typical pulse width value set.
8. The method of claim 7, wherein extracting corresponding pulse widths from a radar pulse train based on the set of real carrier frequency-repetition frequency pairs to obtain a three-parameter feature set of the pulses comprises:
combining the pulse width typical value set with the real carrier frequency-repetition frequency pair set to obtain a three-parameter typical set;
and traversing the carrier frequency-repetition frequency pair typical value set, and respectively counting the pulse width parameters of the radar pulse trains to obtain a complete three-parameter feature set of the radar pulse trains.
9. The method of claim 3, wherein the DBSCAN clustering is adopted, and comprises:
when DBSCAN clustering analysis is carried out on the point set by the parameters, the clustering radius is set to be 1, and the clustering frequency threshold is set according to the pulse number of the electronic reconnaissance radar;
when the DBSCAN clustering analysis is carried out on the pulse pair point set, the clustering radius is set to be 3 times of the pulse width measurement standard deviation, and the clustering frequency threshold is set according to the pulse number of the electronic reconnaissance radar.
10. A radar signal statistical feature extraction device based on electromagnetic data clustering, the device comprising:
a data acquisition module configured to acquire a radar pulse train, the radar pulse train including a plurality of radar pulse sub-trains, the radar pulse sub-trains including a plurality of pulses, the pulses including parameters: carrier frequency, pulse width, arrival time;
the cluster analysis module is used for calculating the inter-pulse time difference of adjacent pulses according to the arrival time to obtain a time difference-carrier frequency pair set of the pulses; determining the parameter pairs with strong aggregation in the time difference-carrier frequency pair set through clustering analysis to obtain a carrier frequency-repeated frequency pair typical value set;
the checking module is used for checking each parameter pair, determining false repetition frequency, and removing the false repetition frequency from the carrier frequency-repetition frequency pair typical value set to obtain a real carrier frequency-repetition frequency pair set;
and the pulse width parameter extraction module is used for extracting corresponding pulse widths from the radar pulse train based on the real carrier frequency-repetition frequency pair set to obtain a three-parameter feature set of the pulses.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210776678.3A CN114839602B (en) | 2022-07-04 | 2022-07-04 | Radar signal statistical feature extraction method and device based on electromagnetic data clustering |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210776678.3A CN114839602B (en) | 2022-07-04 | 2022-07-04 | Radar signal statistical feature extraction method and device based on electromagnetic data clustering |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN114839602A true CN114839602A (en) | 2022-08-02 |
| CN114839602B CN114839602B (en) | 2022-09-16 |
Family
ID=82574142
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210776678.3A Active CN114839602B (en) | 2022-07-04 | 2022-07-04 | Radar signal statistical feature extraction method and device based on electromagnetic data clustering |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114839602B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117554919A (en) * | 2024-01-11 | 2024-02-13 | 成都金支点科技有限公司 | Radar signal sorting and searching method based on bidirectional LSTM network |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US7397415B1 (en) * | 2006-02-02 | 2008-07-08 | Itt Manufacturing Enterprises, Inc. | System and method for detecting and de-interleaving radar emitters |
| US20100283666A1 (en) * | 2009-05-08 | 2010-11-11 | Agency For Defense Development | Radar signals clustering method using frequency modulation characteristics and combination characteristics of signals, and system for receiving and processing radar signals using the same |
| CN103941236A (en) * | 2014-05-12 | 2014-07-23 | 中国船舶重工集团公司第七二四研究所 | Multi-parameter rasterizing sliding signal statistic screening processing method |
| CN106772261A (en) * | 2016-12-07 | 2017-05-31 | 中国船舶重工集团公司第七二四研究所 | Radar detection signal multidimensional characteristic clusters visual display method |
| CN110598738A (en) * | 2019-08-07 | 2019-12-20 | 中国电子科技集团公司第二十九研究所 | Distributed clustering method based on spatial information |
| CN111190146A (en) * | 2020-01-13 | 2020-05-22 | 中国船舶重工集团公司第七二四研究所 | Complex signal sorting method based on visual graphic features |
| CN112986928A (en) * | 2021-03-11 | 2021-06-18 | 哈尔滨工程大学 | Signal sorting multi-source fusion processing method in complex electromagnetic environment |
| CN113721219A (en) * | 2021-10-08 | 2021-11-30 | 中国电子科技集团公司第三十八研究所 | Radar signal sorting method and system based on multi-parameter clustering |
-
2022
- 2022-07-04 CN CN202210776678.3A patent/CN114839602B/en active Active
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US7397415B1 (en) * | 2006-02-02 | 2008-07-08 | Itt Manufacturing Enterprises, Inc. | System and method for detecting and de-interleaving radar emitters |
| US20100283666A1 (en) * | 2009-05-08 | 2010-11-11 | Agency For Defense Development | Radar signals clustering method using frequency modulation characteristics and combination characteristics of signals, and system for receiving and processing radar signals using the same |
| CN103941236A (en) * | 2014-05-12 | 2014-07-23 | 中国船舶重工集团公司第七二四研究所 | Multi-parameter rasterizing sliding signal statistic screening processing method |
| CN106772261A (en) * | 2016-12-07 | 2017-05-31 | 中国船舶重工集团公司第七二四研究所 | Radar detection signal multidimensional characteristic clusters visual display method |
| CN110598738A (en) * | 2019-08-07 | 2019-12-20 | 中国电子科技集团公司第二十九研究所 | Distributed clustering method based on spatial information |
| CN111190146A (en) * | 2020-01-13 | 2020-05-22 | 中国船舶重工集团公司第七二四研究所 | Complex signal sorting method based on visual graphic features |
| CN112986928A (en) * | 2021-03-11 | 2021-06-18 | 哈尔滨工程大学 | Signal sorting multi-source fusion processing method in complex electromagnetic environment |
| CN113721219A (en) * | 2021-10-08 | 2021-11-30 | 中国电子科技集团公司第三十八研究所 | Radar signal sorting method and system based on multi-parameter clustering |
Non-Patent Citations (3)
| Title |
|---|
| MANON MOTTIER 等: "Deinterleaving and Clustering unknown RADAR pulses", 《2021 IEEE RADAR CONFERENCE (RADARCONF21)》 * |
| 刘章孟 等: "多功能雷达脉冲列的语义编码与模型重建", 《雷达学报》 * |
| 吴勇 等: "基于平面分割和PRI的雷达信号分选", 《电子信息对抗技术》 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117554919A (en) * | 2024-01-11 | 2024-02-13 | 成都金支点科技有限公司 | Radar signal sorting and searching method based on bidirectional LSTM network |
| CN117554919B (en) * | 2024-01-11 | 2024-03-29 | 成都金支点科技有限公司 | Radar signal sorting and searching method based on bidirectional LSTM network |
Also Published As
| Publication number | Publication date |
|---|---|
| CN114839602B (en) | 2022-09-16 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US10429488B1 (en) | System and method for geo-locating and detecting source of electromagnetic emissions | |
| KR101007662B1 (en) | Radar Signals Clustering Method using Frequency Modulation Characteristics and Combination Characteristics of Signals, and System for Receiving and Processing Radar Signals using the same | |
| Conte et al. | Statistical analysis of real clutter at different range resolutions | |
| Dudczyk et al. | Specific emitter identification based on graphical representation of the distribution of radar signal parameters | |
| Artyushenko et al. | Statistical characteristics of envelope outliers duration of non-Gaussian information processes | |
| Besic et al. | Unraveling hydrometeor mixtures in polarimetric radar measurements | |
| CN114839602B (en) | Radar signal statistical feature extraction method and device based on electromagnetic data clustering | |
| US20160209492A1 (en) | Method for non-supervised deinterleaving by n-dimensional enrichment | |
| JP6773606B2 (en) | Radar device and its radar signal processing method | |
| CN112926465B (en) | Coastline property identification method and device based on point cloud type | |
| Alves et al. | A statistical analysis for wavelength-resolution SAR image stacks | |
| Campos et al. | False alarm reduction in wavelength–resolution SAR change detection schemes by using a convolutional neural network | |
| US10135488B2 (en) | Method for deinterleaving radar signals | |
| Liu et al. | Unsupervised clustering method to discriminate dense deception jamming for surveillance radar | |
| Wang et al. | A robust constant false alarm rate detector based on the Bayesian estimator for the non-homogeneous Weibull clutter in HFSWR | |
| CN111077532A (en) | Surface feature space information acquisition method based on deconvolution and Gaussian decomposition | |
| Alammar et al. | Automatic extraction of signal areas from radio spectrograms based on the Hough transform | |
| Guillén et al. | Radar detection in the moments space with constant false alarm rate | |
| Jeong et al. | Robust CFAR detector with ordered statistic of sub-reference cells in multiple target situations | |
| Vu et al. | A detector for wavelength resolution SAR incoherent change detection | |
| Yan et al. | Robust Target Localization for Multistatic Passive Radar Networks | |
| CN115598615B (en) | Power spectrum information geometric radar target detection method and device based on sub-band filtering | |
| CN115598616B (en) | Radar target detection method, device and equipment based on power spectrum information geometry | |
| US7103502B1 (en) | Enhanced system for detection of randomness in sparse time series distributions | |
| Liu et al. | Modeling multilook polarimetric SAR images with heavy-tailed rayleigh distribution and novel estimation based on matrix log-cumulants |
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 |