CN107656245B - Method for applying information fusion to radar signal sorting - Google Patents
Method for applying information fusion to radar signal sorting Download PDFInfo
- Publication number
- CN107656245B CN107656245B CN201710722269.4A CN201710722269A CN107656245B CN 107656245 B CN107656245 B CN 107656245B CN 201710722269 A CN201710722269 A CN 201710722269A CN 107656245 B CN107656245 B CN 107656245B
- Authority
- CN
- China
- Prior art keywords
- sorting
- fusion
- data
- pulse
- radar
- 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.)
- Active
Links
- 230000004927 fusion Effects 0.000 title claims abstract description 84
- 238000000034 method Methods 0.000 title claims abstract description 26
- 230000005855 radiation Effects 0.000 claims abstract description 20
- 238000012163 sequencing technique Methods 0.000 claims abstract description 6
- 238000005259 measurement Methods 0.000 claims description 14
- 230000009471 action Effects 0.000 claims description 6
- 238000000926 separation method Methods 0.000 claims description 6
- 230000008569 process Effects 0.000 claims description 4
- 238000001514 detection method Methods 0.000 claims description 3
- 238000013507 mapping Methods 0.000 claims description 3
- 238000004088 simulation Methods 0.000 abstract description 6
- 238000012545 processing Methods 0.000 abstract description 2
- 230000000007 visual effect Effects 0.000 abstract description 2
- 238000013528 artificial neural network Methods 0.000 description 3
- 230000000694 effects Effects 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 2
- 230000007613 environmental effect Effects 0.000 description 2
- 238000007499 fusion processing Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000007500 overflow downdraw method Methods 0.000 description 2
- 238000012360 testing method Methods 0.000 description 2
- 241000764238 Isis Species 0.000 description 1
- 238000009825 accumulation Methods 0.000 description 1
- 239000008186 active pharmaceutical agent Substances 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000015572 biosynthetic process Effects 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 230000001934 delay Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 239000000203 mixture Substances 0.000 description 1
- 230000003252 repetitive effect Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 230000001360 synchronised effect Effects 0.000 description 1
- 238000003786 synthesis reaction Methods 0.000 description 1
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
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Radar, Positioning & Navigation (AREA)
- Remote Sensing (AREA)
- Medicines Containing Antibodies Or Antigens For Use As Internal Diagnostic Agents (AREA)
- Radar Systems Or Details Thereof (AREA)
Abstract
The invention discloses a method for applying information fusion to radar signal sorting, and belongs to the field of radar signal processing. The invention performs data-level fusion on Pulse Description Words (PDW) before radar signal sorting, and performs feature-level fusion on sorting results after sorting. A time alignment method based on first-level pulse arrival Time (TOA) difference matching is proposed in data-level fusion, and D-S evidence theoretical judgment correlation is applied to pulses arriving at the same time to obtain a new PDW; and unifying parameters describing the same radar in the feature level fusion and sequencing the sorting results with credibility. The method solves the problems that the single receiving device possibly fails to receive the pulse and the same radar description parameter is not completely the same after the pulse is selected. Simulation results show that the success rate of radar signal sorting can be effectively improved through data level fusion, and more concise and visual radiation source information can be obtained through application of feature level fusion.
Description
Technical Field
The invention belongs to the field of radar signal processing, and particularly relates to a method for applying information fusion to radar signal sorting.
Background
With the development of science and technology, the electromagnetic signal environment of modern battlefields is increasingly complex, the radiation source density is multiplied, and the electromagnetic signal patterns are more complex and changeable. Modern radars have multimode and concealment, one radar often has multiple working modes to realize different functions, each mode also has corresponding technical characteristic parameters, and the application of radar anti-interference measures such as low interception probability causes that partial characteristic parameters obtained by reconnaissance receiving equipment have defects, have incomplete and uncertain characteristics, and bring great difficulty to radar signal sorting. Considering that the information of each reconnaissance receiving device has complementarity in a certain observation time, the Pulse Description Word (PDW) received by the receiver can be subjected to fusion processing and then sorted by an information fusion method, and a better sorting effect is obtained compared with that of a single receiving device. The continuous observation and sorting result of the same reconnaissance receiving device has the characteristic of repeated appearance, but is influenced by factors such as sorting algorithm, measurement accuracy and the like, parameters describing the same radar are possibly not identical, all the parameters are unified through the fusion of the sorting results, and the sorted radars are subjected to credibility sorting, so that radiation source information in observation time can be reflected simply and visually.
At present, scholars at home and abroad mainly aim at carrying out fusion recognition on a radiation source after separation aiming at the research direction of information fusion and radar signal separation. In 2015, Wuzheng et al effectively combine a probabilistic neural network, a multivariate attribute fusion algorithm and a D-S evidence theory to provide an information fusion model for radar radiation source identification, wherein the model firstly identifies radar parameters obtained by radar reconnaissance equipment by using the probabilistic neural network, and then identifies possible radar models identified by the neural network by using the multivariate information fusion algorithm when the identification result is not unique, so as to finally determine the identification result. The fusion of the pulse description words before sorting is limited by documents, and related articles can not be searched, but if the data before sorting is fused, more battlefield information can be reserved, and the discovery probability and the recognition capability of targets can be effectively improved.
Disclosure of Invention
The invention aims to provide a method for applying information fusion to radar signal sorting, which is used for applying information fusion to radar signal sorting to achieve better sorting effect than single equipment and obtain simpler and more visual radiation source information.
The purpose of the invention is realized by the following technical scheme:
the invention firstly carries out data level fusion on PDW obtained by different receivers, and then carries out fusion of sorting results on pulse data obtained by the same receiver in continuous observation time, namely, carries out feature level fusion on the sorting results, so that the application of information fusion in radar signal sorting comprises two parts of data level PDW fusion and feature level sorting result fusion.
1. A method for applying information fusion to radar signal classification is characterized by comprising two parts of data-level PDW fusion and feature-level classification result fusion, wherein:
the data-level PDW fusion mainly comprises the following steps:
(1) time alignment, i.e. registering the unsynchronized data from the receivers to the same time;
(1.1) carrying out primary TOA difference matching on two groups of observation data in the same time period, and calculating to obtain a primary TOA difference of the observation data;
(1.2) adding etoa1And etoa2Sequentially forming a vector X of dimension l1(i)、X2(j)，I.e. respectively selecting a pulse stream consisting of l pulses for matching
(1.3) statistics of X1(i) And X2(j) The number k of corresponding errors between the measurement errors and the error and error of the l-dimensional data are within the measurement errors;
(1.4) calculating a first-order TOA difference similarity rate;
(1.5) setting a corresponding error minimum value (min { er (i, j) }) as an optimal matching position, and defining the relative time of the first group in an error range as the relative time delta t of the two groups of observation data;
(2) attribute correlation, namely calculating carrier frequency and pulse width support for the pulse which arrives at the same time after time alignment;
mapping the difference degree of the carrier frequency f and the pulse width pw to an interval of [0,1] through a probability assignment function, and calculating the support degree of the two parameters;
(3) calculating the total support degree by using a D-S fusion formula, and setting a threshold to judge whether the arriving pulses at the same time come from the same radiation source;
D-S evidence theory uses a recognition framework U to represent a proposition set of interest, and defines a basic probability assignment function m:2 on UU→[0,1]And satisfies the following conditions:
m (-) is called the fundamental probability assignment on U, if m (A) ≠ 0, then A is called a focal element, m (A) reflects the confidence level of A;
for an inference system with two features, the probability assignments for the two features are m1、m2For the subset A, the overall basic probability assignment under the combined action of the two characteristics is obtained
D-S rule of (1):
wherein
If the conflict item is a conflict item, if K is not equal to 1, determining m basic probability assignments; if K is 1, then m is considered to be1And m2Contradictory, no combination is necessary.
(4) Threshold judgment, fusing different data from the same radiation source at the same time into a new data value, arranging the data into a new PDW according to the sequence of arrival time,
if the signals come from the same radiation source, calculating the arithmetic mean value of the signals as the carrier frequency and pulse width value of the pulse at the moment after fusion, and sequencing the pulses which do not arrive at the same time according to the sequence of the arrival time to obtain more comprehensive information;
wherein, aiming at the condition of pulse advance, the arrival time value after fusion is defined as the smaller value of the known two arrival times, and the pulse width value is the difference value between the pulse back edge with larger value and the newly defined arrival time; aiming at the condition that the pulse contains, defining the arrival time and the pulse width value of the fused pulse as parameters corresponding to the contained pulse; if the above phenomenon does not exist, the data are arranged according to the arrival time sequence.
The fusion of the feature fraction sorting results mainly comprises the following steps:
the support degrees of three attributes of carrier frequency, pulse width and PRI are calculated, then the D-S evidence theory combination rule is adopted to carry out attribute fusion to obtain the total support degree, and all characteristic parameters describing the same radar are weighted to obtain uniform parameters.
Calculating carrier frequency f support degree m of radarfAnd pulse width support mpw；
PRI is divided into two types, fixed and ragged:
A. the sorting results were for conventional radar with a fixed PRI:
one sort result for a set of observation time periods is TaOne sorting result of another set of observation periods is TbThen the difference between the two sorting results is Δ Tab＝|Ta-TbI, PRI support m for two sorting resultsPRIComprises the following steps:
Tis the error of the repetition period of the reconnaissance system;
B. the separation result is a spread radar:
taking two spread as an example, the skeleton period of a group of observed results is TaWith a repeating interval of two-step difference of Ta1And Ta2The other component of the selection result shows the framework period TbWith a repeating interval of two-step difference of Tb1And Tb2(ii) a PRI support m of sorting results at this timePRIComprises the following steps:
wherein the difference Δ T ═ T of the PRIa-Tb|，ΔTabi＝|Tai-Tbi|，i＝1,2；
The total probability assignment function under the combined action of n evidences is expressed as follows:
if the total support degree is greater than the set threshold value, the two sorting results which are compared describe the same radar, the results obtained after sorting all the detection data in the observation time are subjected to feature level fusion in sequence, and if the results are described as the same radar, the reciprocal of the PRI variance of the sorting results is used as a weight coefficient, so that the feature parameters of the same radar are unified;
in order to avoid sorting errors caused by a sorting algorithm, for example, a PRI parameter displayed by sorting is a multiple of an actual PRI, the pulse number and frequency number of the radar are required to be described in a statistical manner in the process of realizing parameter unification, the frequency of the sorting result is high, the number of used pulses is large, the reliability of the sorting result is considered to be high, and radiation source information in an observation time period can be intuitively obtained by performing statistical sequencing on the pulse numbers describing different radars.
The invention has the beneficial effects that:
for data received by several receivers observing the same target, direct sorting of the data using a single receiving device may result in sorting failures due to environmental, noise, and possible loss of pulses. The invention firstly carries out data level fusion aiming at PDWs obtained by different receivers, thus solving the problem of possible sorting failure after the pulse received by a single receiving device is lost;
and then, fusion of sorting results is carried out on pulse data obtained by the same receiver in continuous observation time, namely, feature level fusion is carried out on the sorting results, so that the problem that the description parameters of the same radar are not completely the same after sorting is solved.
Drawings
FIG. 1 is a PDW fusion model;
FIG. 2 is a feature level selection result fusion model;
FIG. 3 is a schematic diagram of PDW time alignment;
FIG. 4 is a fusion rule;
FIG. 5 shows the sorting results of lost signals;
fig. 6 shows the signal sorting results after fusion.
Detailed Description
The following further describes embodiments of the present invention with reference to the accompanying drawings:
a method for applying information fusion to radar signal classification specifically comprises data-level PDW fusion and feature-level classification result fusion, and models are shown in figures 1 and 2.
The data-level PDW fusion comprises the following steps:
the method comprises the following steps: and (4) time alignment. Because the receivers placed at different positions receive radar signals independently, different delays of a communication network may exist, and therefore, time differences may exist in the reporting time of each receiver. Before fusion, data from multiple receivers needs to be transformed in the same time frame of reference, and information which is not synchronous needs to be registered to the same fusion time. The alignment principle is shown in fig. 3, and the specific method is as follows:
1. and performing primary TOA difference matching on the two groups of observation data in the same time period, and calculating to obtain the primary TOA difference of the observation data.
etoa1(i)＝toa1(i+1)-toa1(i),i＝1,2…n1-1 (1)
etoa2(j)＝toa2(j+1)-toa2(j),j＝1,2…n2-1 (2)
In the formula n1And n2The number of pulses of the observed data, n1-1 and n2-1 are the number of pulses of the first-order TOA difference to be matched, respectively.
2. Will etoa1And etoa2Sequentially forming a vector X of dimension l1(i)、X2(j) In that respect I.e. a pulse stream consisting of l pulses is selected for matching.
X1(i)＝[etoa1(i),etoa1(i+1)…etoa1(i+l-1)],i＝1,2…(n1-1)-l+1 (3)
X2(j)＝[etoa2(j),etoa2(j+1)…etoa2(j+l-1)],j＝1,2…(n2-1)-l+1 (4)
3. Statistics of X1(i) And X2(j) The number k of corresponding errors between the two and the l-dimensional data corresponds to the error and error.
error(i,j)＝|X1(i)-X2(j)|,i＝1,2…n1-l；j＝1,2…n2-l (5)
Calculating the first-order TOA difference similarity rate:
4. sorting according to the number k in the error, and taking the position with k as the maximum value as the best matching position; if the number of k being the maximum value is not 1, the correspondence error is compared with error. And setting the corresponding error minimum value (min { er (i, j) }) as the best matching position, and defining the relative time of the first group in the error range as the relative time delta t of the two groups of observed data.
Δt＝toa1(i)-toa2(j) (7)
i. j is the best match position.
Step two: and (4) attribute correlation, namely calculating carrier frequency and pulse width support degree of the pulse which arrives at the same time after time alignment. Mapping the difference degree of the carrier frequency f and the pulse width pw to [0,1] through a probability assignment function]The support degree of the two parameters is calculated. If a signal is received by both receivers simultaneously, the measurement of a set of observations is f for a certain property parameter, e.g. carrier frequency faAnother set of measurements is fbThe difference between these two measurements is then Δ fab＝|fa-fbL. At this time, the similarity m between the two measured valuesfComprises the following steps:
wherein f isIs the measurement tolerance of the sensor, the measurement accuracy is chosen to be 4 times.
The pulse width support m of two measured values is obtained in the same waypw：
Step three: and (4) solving the overall support degree by using a D-S fusion formula, and setting a threshold to judge whether the arriving pulses at the same time come from the same radiation source. The measurement of the basic parameters of the target signal provided by the reconnaissance receiving device depends on the signal pattern radiated by the target radiation source, and the loss or deviation of the signal measurement parameters can be caused due to the influence of environmental noise, multipath and other factors. The D-S evidence theory is the popularization of the Bayesian reasoning method, and can better describe the uncertain and unknown type problems, so that the DS evidence theory formula can be adopted to perform fusion processing on the PDW parameters received by different receivers.
D-S evidence theory uses the recognition framework U to express what feelsThe proposition set of interest defines the basic probability assignment function m:2 on UU→[0,1]And satisfies the following conditions:
m (-) is called the basic probability assignment over U. If m (A) ≠ 0, A is called a focus, and m (A) reflects the reliability of A.
The evidence theory method has own reasonable synthesis rule, which is called D-S rule for short. For an inference system with two features, the probability assignments for the two features are m1、m2For the subset A, the overall basic probability assignment under the combined action of the two characteristics is obtained
The D-S rule of (A) is as shown in formula (11).
Wherein
Is a conflicting item. If K is not equal to 1, determining m basic probability assignments; if K is 1, then m is considered to be1And m2Contradictory, no combination is necessary.
Step four: and (4) threshold judgment, fusing different data from the same radiation source at the same time into a new data value, and arranging the data into a new PDW according to the sequence of arrival time.
By a time alignment method based on primary TOA difference matching, asynchronous information is registered to the same fusion time, the relevance of signals received by different receivers at the same time can be calculated by a data correlation technology based on a D-S evidence theory, and the signals can be judged whether come from the same radiation source or not by comparing with a threshold of 0.5. If the signals come from the same radiation source, the arithmetic mean value is calculated as the carrier frequency and pulse width value of the pulse at the moment after fusion. For the pulses which do not arrive at the same time, the pulses are sequenced according to the sequence of the arrival time, and more comprehensive information is obtained. The specific fusion rule is shown in fig. 4.
1) For the case of pulse advance, the arrival time value after fusion is defined as the smaller value of the known two arrival times, and the pulse width value is the difference value between the pulse trailing edge with the larger value and the newly defined arrival time.
2) For the situation of pulse inclusion, the arrival time and the pulse width value of the fused pulse are defined as the parameters corresponding to the included pulse.
3) If the above phenomenon does not exist, the data are arranged according to the arrival time sequence.
The fusion method of the feature level sorting results comprises the following steps:
for the sorting results of the same receiver in different observation time periods, the overall support degree can be obtained by calculating the support degrees of three attributes of the carrier frequency, the pulse width and the PRI of the receiver and adopting a combination rule based on a D-S evidence theory to perform attribute fusion on the characteristic parameters of the radiation source.
The similarity of the properties of the carrier frequency and the pulse width of the radar is the same as that of the formulas (8) and (9), and the PRI is divided into a fixed type and a staggered type.
1) The sorting results were for conventional radar with a fixed PRI:
one sort result for a set of observation time periods is TaOne sorting result of another set of observation periods is TbThen the difference between the two sorting results is Δ Tab＝|Ta-TbL. PRI similarity m of two sorting resultsPRIComprises the following steps:
Tto detect the repetitive cycle error of the system.
2) The separation result is a spread radar:
taking two spread as an example, the skeleton period of a group of observed results is TaWith a repeating interval of two-step difference of Ta1And Ta2The other component of the selection result shows the framework period TbWith a repeating interval of two-step difference of Tb1And Tb2. PRI similarity m of sorting results at this timePRIComprises the following steps:
wherein the difference Δ T ═ T of the PRIa-Tb|，ΔTabi＝|Tai-Tbi|，i＝1,2。
The D-S rule of equation (11) is for an equation with only two evidences, carrier frequency and pulse width, and we obtain a new parameter PRI after sorting. The total probability assignment function under the combined action of n evidences is expressed as follows:
and if the overall support degree is greater than the set threshold value, judging that the two sorting results for comparison describe the same radar. And sequentially performing feature level fusion on results obtained after sorting all the detection data in the observation time, and if the results are described by the same radar, taking the inverse PRI variance of the sorted results as a weight coefficient to realize the unification of the feature parameters of the same radar.
In order to avoid sorting errors caused by a sorting algorithm, for example, the PRI parameter displayed by sorting is a multiple of the actual PRI, the pulse number and frequency number of the radar are required to be described in a statistical manner in the process of realizing parameter unification. The sorting result has a high frequency of occurrence and a large number of pulses are used, and the reliability of the sorting result is considered to be high. Through counting and sequencing the pulse numbers describing different radars, the radiation source information in the observation time period can be intuitively obtained.
Example (b):
example 1: supposing that two reconnaissance equipment receivers work simultaneously and the information obtained by each reconnaissance receiving equipment is different from each other, the PDWs received by the receivers are fused by a data level fusion algorithm to form a new PDW, and the PDWs before and after fusion are respectively sorted by a conventional SDIF method. The simulation constructs two pulse descriptors of a mixture of conventional radar signals, whose parameters are shown in table 1 below.
TABLE 1 Radar simulation parameter set-up
The missing signal was compared to the fused signal sorting results at a pulse missing rate of 50%, as shown in table 2.
TABLE 2 data-level Pre-and post-fusion sorting results
As can be seen from table 2, if the received signals are missing as shown by the first path and the second path of signals in the table, radars with PRI parameters of 400us and 500us cannot be correctly sorted, and two radars can be successfully sorted by obtaining more pulse numbers after data level fusion.
Example 2: the test signal intercepts the influence of the loss rate on the signal sorting. And conventional sorting is carried out on the conventional radar signals and the signals subjected to data level fusion under different loss rates by applying the simulation parameters in the table 1. The results of sorting the single-pass signal with missing pulses and after data level fusion are shown in fig. 5 and 6.
As can be seen from fig. 5, in the process of the pulse loss rate of 10% to 45%, two paths of lost signals can be successfully selected and set for two radars; when the pulse loss rate reaches 50%, the lost two-path signals can only correctly sort out the conventional radar with the PRI parameter of 500us, and cannot correctly sort out the radar with the PRI parameter of 400 us. As can be seen from FIG. 6, after data-level fusion, the sorting can still be successfully performed when the pulse loss rate reaches 60%, and the effectiveness of the radar signal data-level fusion algorithm in improving the sorting success rate is verified.
Example 3: and testing the fusion effect of the characteristic grade separation results. Simulation firstly carries out sorting on PDW data packets reported after a certain reconnaissance receiver continuously observes for 5 times, and sorting results are shown in a table 3. And selecting data in simulation as actually measured data, wherein the selection algorithm still adopts an SDIF method.
TABLE 3 sorting results under multiple observations
The results obtained by performing feature level fusion on a plurality of sorting results obtained in table 3 are shown in table 4. It can be seen that although the pulse number of the radar describing the radar with the PRI of 1.9613ms is only 22 in the 76 pulses detected by the receiver in the observation period with the time number of 1, and even the receiver does not detect the pulse signal describing the radar in the observation period with the time number of 5, through multiple observation and sorting accumulation, the possibility of sorting failure can be eliminated, and it is determined that the conventional radar with the PRI of 1.9613ms does exist.
TABLE 4 feature level fusion results
The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention, and various modifications and changes may be made by those skilled in the art. Any modification, equivalent replacement, or improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.
Claims (6)
1. A method for applying information fusion to radar signal classification is characterized by comprising two parts of data-level PDW fusion and feature-level classification result fusion, wherein:
the data-level PDW fusion mainly comprises the following steps:
(1) time alignment, i.e. registering the unsynchronized data from the receivers to the same time;
(2) attribute correlation, namely calculating carrier frequency and pulse width support for the pulse which arrives at the same time after time alignment;
(3) calculating the total support degree by using a D-S fusion formula, and setting a threshold to judge whether the arriving pulses at the same time come from the same radiation source;
(4) threshold judgment, fusing different data from the same radiation source at the same time into a new data value, arranging the data into a new PDW according to the sequence of arrival time,
the fusion of the feature fraction sorting results mainly comprises the following steps:
the support degrees of three attributes of carrier frequency, pulse width and PRI are calculated, then the D-S evidence theory combination rule is adopted to carry out attribute fusion to obtain the total support degree, and all characteristic parameters describing the same radar are weighted to obtain uniform parameters.
2. The method for applying information fusion to radar signal sorting according to claim 1, wherein the step (1) of the data-level PDW fusion is specifically:
(1.1) carrying out primary TOA difference matching on two groups of observation data in the same time period, and calculating to obtain the primary TOA difference of the observation data:
etoa1(i)＝toa1(i+1)-toa1(i),i＝1,2…n1-1
etoa2(j)＝toa2(j+1)-toa2(j),j＝1,2…n2-1
in the formula n1And n2The number of pulses of the observed data, n1-1 and n2-1 is the number of pulses of the first-order TOA difference to be matched, respectively;
(1.2) adding etoa1And etoa2Sequentially forming a vector X of dimension l1(i)、X2(j) I.e. consisting of one pulse eachPulse flow matching:
X1(i)＝[etoa1(i),etoa1(i+1)…etoa1(i+l-1)],i＝1,2…(n1-1)-l+1
X2(j)＝[etoa2(j),etoa2(j+1)…etoa2(j+l-1)],j＝1,2…(n2-1)-l+1
(1.3) statistics of X1(i) And X2(j) The number k of corresponding errors between the two and the corresponding error of the l-dimensional data and error:
error(i,j)＝|X1(i)-X2(j)|,i＝1,2…n1-l；j＝1,2…n2-l
(1.4) calculating the first-order TOA difference similarity rate:
sorting according to the number k in the error, and taking the position with k as the maximum value as the best matching position; if the number of k as the maximum value is not 1, comparing the corresponding error with error;
(1.5) setting the corresponding error minimum value (min { er (i, j) }) as the best matching position, and defining the relative time of the first group in the error range as the relative time delta t of the two groups of observation data:
Δt＝toa1(i)-toa2(j)
where i, j are the best matching positions.
3. The method for applying information fusion to radar signal sorting according to claim 1, wherein the step (2) of the data-level PDW fusion specifically comprises:
mapping the difference degree of the carrier frequency f and the pulse width pw to an interval of [0,1] through a probability assignment function, and calculating the support degree of the two parameters;
the carrier frequency f of a set of observed data is measured as faAnother set of measurements is fbThe difference between these two measurements is then Δ fab＝|fa-fb|；
Support m between these two carrier frequency f measured valuesfComprises the following steps:
wherein f isThe measurement tolerance of the sensor is selected to be 4 times of measurement precision;
the pulse width support m of two measured values is obtained in the same waypw：
4. The method for applying information fusion to radar signal sorting according to claim 1, wherein the step (3) of the data-level PDW fusion is specifically:
D-S evidence theory uses a recognition framework U to represent a proposition set of interest, and defines a basic probability assignment function m:2 on UU→[0,1]And satisfies the following conditions:
m (-) is called the fundamental probability assignment on U, if m (A) ≠ 0, then A is called a focal element, m (A) reflects the confidence level of A;
for an inference system with two features, the probability assignments for the two features are m1、m2For the subset A, the overall basic probability assignment under the combined action of the two characteristics is solved
D-S rule of (1):
wherein
If the conflict item is a conflict item, if K is not equal to 1, determining m basic probability assignments; if K is 1, then m is considered to be1And m2Contradictory, no combination is necessary.
5. The method for applying information fusion to radar signal sorting according to claim 1, wherein the step (4) of the data-level PDW fusion is specifically:
if the signals are pulses arriving at the same time, calculating the arithmetic mean value of the signals as the carrier frequency and pulse width value of the pulses at the moment after fusion, and sequencing the pulses which do not arrive at the same time according to the sequence of arrival time to obtain more comprehensive information;
wherein,
for the condition of pulse advance, defining the arrival time value after fusion as the smaller value of the known two arrival times, and defining the pulse width value as the difference value of the pulse back edge with larger value and the newly defined arrival time;
aiming at the condition that the pulse contains, defining the arrival time and the pulse width value of the fused pulse as parameters corresponding to the contained pulse;
if the above phenomenon does not exist, the data are arranged according to the arrival time sequence.
6. The method according to claim 1, wherein the fusion of the feature-level classification results is specifically:
calculating carrier frequency f support degree m of radarfAnd pulse width support mpw；
PRI is divided into two types, fixed and ragged:
A. the sorting results were for conventional radar with a fixed PRI:
one sort result for a set of observation time periods is TaOne sorting result of another set of observation periods is TbThen the difference between the two sorting results is Δ Tab＝|Ta-TbI, PRI support m for two sorting resultsPRIComprises the following steps:
Tis the error of the repetition period of the reconnaissance system;
B. the separation result is a spread radar:
taking two spread as an example, the skeleton period of a group of observed results is TaWith a repeating interval of two-step difference of Ta1And Ta2The other component of the selection result shows the framework period TbWith a repeating interval of two-step difference of Tb1And Tb2(ii) a PRI support m of sorting results at this timePRIComprises the following steps:
wherein the difference Δ T ═ T of the PRIa-Tb|，ΔTabi＝|Tai-Tbi|，i＝1,2；
The total probability assignment function under the combined action of n evidences is expressed as follows:
if the total support degree is greater than the set threshold value, the two sorting results which are compared describe the same radar, the results obtained after sorting all the detection data in the observation time are subjected to feature level fusion in sequence, and if the results are described as the same radar, the reciprocal of the PRI variance of the sorting results is used as a weight coefficient, so that the feature parameters of the same radar are unified;
in order to avoid sorting errors caused by a sorting algorithm, for example, a PRI parameter displayed by sorting is a multiple of an actual PRI, the pulse number and frequency number of the radar are required to be described in a statistical manner in the process of realizing parameter unification, the frequency of the sorting result is high, the number of used pulses is large, the reliability of the sorting result is considered to be high, and radiation source information in an observation time period can be intuitively obtained by performing statistical sequencing on the pulse numbers describing different radars.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710722269.4A CN107656245B (en) | 2017-08-22 | 2017-08-22 | Method for applying information fusion to radar signal sorting |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710722269.4A CN107656245B (en) | 2017-08-22 | 2017-08-22 | Method for applying information fusion to radar signal sorting |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN107656245A CN107656245A (en) | 2018-02-02 |
| CN107656245B true CN107656245B (en) | 2020-12-04 |
Family
ID=61128457
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710722269.4A Active CN107656245B (en) | 2017-08-22 | 2017-08-22 | Method for applying information fusion to radar signal sorting |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107656245B (en) |
Families Citing this family (23)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108562875B (en) * | 2018-04-24 | 2022-03-15 | 扬州健行电子科技有限公司 | Method for sorting radar signals by priori database matching method |
| CN109270497B (en) * | 2018-10-28 | 2023-03-31 | 西南电子技术研究所（中国电子科技集团公司第十研究所） | Multidimensional parameter pre-sorting method for radar pulse signals |
| CN109709555B (en) * | 2018-12-13 | 2020-09-11 | 中国气象局气象探测中心 | Method and system for identifying difference of adjacent scan data of weather radar |
| CN109683143A (en) * | 2019-03-07 | 2019-04-26 | 西安电子科技大学 | Radar Signal Sorting Method, device, computer equipment and can storage medium |
| CN110880012B (en) * | 2019-09-29 | 2023-09-05 | 扬州船用电子仪器研究所(中国船舶重工集团公司第七二三研究所) | Inter-pulse agile radar radiation source frequency information correlation method for multi-reconnaissance platform |
| CN110907908B (en) * | 2019-10-22 | 2023-06-09 | 扬州船用电子仪器研究所(中国船舶重工集团公司第七二三研究所) | Navigation radar signal sorting method based on envelope analysis |
| CN110687511B (en) * | 2019-11-14 | 2023-01-03 | 扬州船用电子仪器研究所(中国船舶重工集团公司第七二三研究所) | Pulse timing sequence recovery method of radar signal simulator |
| CN110809005B (en) * | 2019-11-14 | 2022-05-20 | 扬州船用电子仪器研究所(中国船舶重工集团公司第七二三研究所) | Unified characterization method for radar pulse description words |
| CN110927679B (en) * | 2019-11-14 | 2023-06-09 | 扬州船用电子仪器研究所(中国船舶重工集团公司第七二三研究所) | Hash table-based multi-part radar signal ordering method |
| CN111090077B (en) * | 2020-01-20 | 2023-01-20 | 上海交通大学 | Method for sorting TOA differential statistical radar signals after multi-dimensional de-interleaving |
| CN111447019B (en) * | 2020-03-05 | 2021-12-31 | 中国电子科技集团公司第二十九研究所 | Device for fusing pulse signals among multiple modules |
| CN111563532B (en) * | 2020-04-07 | 2022-03-15 | 西北工业大学 | Unknown target identification method based on attribute weight fusion |
| CN111487584B (en) * | 2020-04-20 | 2022-10-14 | 中国航天科工集团八五一一研究所 | Pulse screening method based on time slice |
| CN111751792B (en) * | 2020-06-15 | 2024-03-22 | 哈尔滨工程大学 | PDW preprocessing method for passive radar signal sorting |
| CN112034426B (en) * | 2020-08-27 | 2024-04-12 | 上海朱光亚战略科技研究院 | Radar signal processing method, device, computer equipment and storage medium |
| CN112231116B (en) * | 2020-10-12 | 2024-06-25 | 航天科工广信智能技术有限公司 | Object fusion method of microwave radar and application system thereof |
| CN112597820A (en) * | 2020-12-10 | 2021-04-02 | 南京长峰航天电子科技有限公司 | Target clustering method based on radar signal sorting |
| CN112986928B (en) * | 2021-03-11 | 2022-06-17 | 哈尔滨工程大学 | Signal sorting multi-source fusion processing method in complex electromagnetic environment |
| CN113075637B (en) * | 2021-04-29 | 2022-06-07 | 中国船舶重工集团公司第七二三研究所 | Airborne PD radar signal sorting method based on pulse descriptor data compression |
| CN113311394B (en) * | 2021-05-18 | 2024-08-23 | 中国船舶重工集团公司第七二三研究所 | Ultra-wideband PDW real-time fusion method |
| CN113625242B (en) * | 2021-07-23 | 2023-09-29 | 哈尔滨工程大学 | Radar signal sorting method based on potential distance graph combined PCA and improved cloud model |
| CN115047417B (en) * | 2022-08-15 | 2022-10-25 | 成都锐新科技有限公司 | Method and system for simulating radar echo and interference |
| CN117347961B (en) * | 2023-12-04 | 2024-02-13 | 中国电子科技集团公司第二十九研究所 | Radar function attribute identification method based on Bayesian learning |
Family Cites Families (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101697006B (en) * | 2009-09-18 | 2011-11-16 | 北京航空航天大学 | Target identification method based on data fusion of airborne radar and infrared imaging sensor |
| CN101893704B (en) * | 2010-07-20 | 2012-07-25 | 哈尔滨工业大学 | Rough set-based radar radiation source signal identification method |
| US9499185B2 (en) * | 2013-12-20 | 2016-11-22 | Thales Canada Inc | Wayside guideway vehicle detection and switch deadlocking system with a multimodal guideway vehicle sensor |
| CN105046067B (en) * | 2015-07-03 | 2017-12-29 | 西北工业大学 | Multiple sensor information amalgamation method based on evidence similarity |
| CN106093873B (en) * | 2016-06-04 | 2018-08-31 | 陈昌孝 | A kind of radar emitter signal recognition methods based on multisensor |
| CN106778907A (en) * | 2017-01-11 | 2017-05-31 | 张军 | A kind of intelligent travelling crane early warning system based on multi-sensor information fusion |
-
2017
- 2017-08-22 CN CN201710722269.4A patent/CN107656245B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN107656245A (en) | 2018-02-02 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107656245B (en) | Method for applying information fusion to radar signal sorting | |
| Kawalec et al. | Specific emitter identification using intrapulse data | |
| CN104931960B (en) | Trend message and radar target situation information full flight path segment data correlating method | |
| CN105354356A (en) | Radar intelligence simulation based air intelligence fusion performance evaluation system and method | |
| Gupta et al. | Electronic warfare: Issues and challenges for emitter classification | |
| CN109002835A (en) | A kind of particle filter data correlation method based on maximum entropy fuzzy clustering | |
| CN112036239B (en) | Radar signal working mode identification method and system based on deep learning network | |
| CN108562893A (en) | A kind of external illuminators-based radar multistation combined tracking method | |
| Miao et al. | Application of one-class classification in NLOS identification of UWB positioning | |
| CN112990125B (en) | Method for judging whether radiation source radar belongs to target platform | |
| CN116340533B (en) | Satellite-borne electromagnetic spectrum big data intelligent processing system based on knowledge graph | |
| Dudczyk et al. | Applying the distance and similarity functions to radar signals identification | |
| CN106093891A (en) | Radar fence anti-crowding measure false target jamming profile method based on doppler velocity inspection | |
| CN110472188A (en) | A kind of abnormal patterns detection method of facing sensing data | |
| CN106019250A (en) | Repeating false target discriminating method based on angular glint | |
| CN113075620B (en) | Signal sorting method based on multi-station time difference grid clustering | |
| Li et al. | Signals Deinterleaving for ES systems using Improved CFSFDP Algorithm | |
| Zhao et al. | Calibrating the transmitter and receiver location errors for moving target localization in multistatic passive radar | |
| CN110880012B (en) | Inter-pulse agile radar radiation source frequency information correlation method for multi-reconnaissance platform | |
| CN114839607A (en) | MFR waveform unit extraction method based on step-by-step variable threshold isolated forest | |
| Adibi et al. | Measuring confidence intervals in link discovery: a bootstrap approach | |
| CN115017958A (en) | Non-line-of-sight signal identification method based on channel characteristic weighting model | |
| Han et al. | Radar clutter classification using expectation-maximization method | |
| Cheng et al. | A radar main lobe pulse correlation sorting method | |
| CN112763989B (en) | CDIF-based jitter signal sorting method |
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 |