CN105223557B - Airborne early warning radar clutter suppression method based on accessory channel - Google Patents

Airborne early warning radar clutter suppression method based on accessory channel Download PDF

Info

Publication number
CN105223557B
CN105223557B CN201510717941.1A CN201510717941A CN105223557B CN 105223557 B CN105223557 B CN 105223557B CN 201510717941 A CN201510717941 A CN 201510717941A CN 105223557 B CN105223557 B CN 105223557B
Authority
CN
China
Prior art keywords
num
early warning
accessory channel
warning radar
dimensionality reduction
Prior art date
Application number
CN201510717941.1A
Other languages
Chinese (zh)
Other versions
CN105223557A (en
Inventor
王彤
张莹莹
蔡启程
吴建新
Original Assignee
西安电子科技大学
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 西安电子科技大学 filed Critical 西安电子科技大学
Priority to CN201510717941.1A priority Critical patent/CN105223557B/en
Publication of CN105223557A publication Critical patent/CN105223557A/en
Application granted granted Critical
Publication of CN105223557B publication Critical patent/CN105223557B/en

Links

Abstract

The invention discloses a kind of airborne early warning radar clutter suppression method based on accessory channel, its thinking is:The three-dimensional echo data that airborne early warning radar receives is obtained, and obtains the clutter ridge of the three-dimensional echo data accordingly, and then obtains two-dimentional echo data;Further according to the clutter ridge of three-dimensional echo data, num accessory channel and num search passage are obtained respectively, obtain temporal frequency vector corresponding to spatial domain frequency vector corresponding to num accessory channel and num accessory channel, and then matrix corresponding to dimensionality reduction matrix corresponding to num accessory channel and num search passage is obtained successively, and the dimensionality reduction matrix for optimizing accessory channel and the transformation matrix based on accessory channel are obtained accordingly;Dimension-reduction treatment is carried out respectively to matrix corresponding to the two-dimentional echo data and num search passage, obtain the search passage steering vector after the echo data after dimensionality reduction and dimensionality reduction, and then obtain the filtered vector after dimensionality reduction, and clutter recognition processing is carried out to the matrix of num search passage accordingly, obtain range Doppler figure.

Description

Airborne early warning radar clutter suppression method based on accessory channel

Technical field

The invention belongs to radar clutter suppression technology field, more particularly to a kind of airborne early warning radar based on accessory channel Clutter suppression method, suitable for solving the problems, such as the decline of airborne early warning radar clutter rejection non-homogeneous clutter environment, change It is apt to its clutter recognition performance.

Background technology

Airborne early warning radar can flexibly, rapidly be deployed in required local and in widespread attention, its master with it It is that target is detected in clutter background to want task, and the target to detecting carries out locating and tracking, and clutter is effectively pressed down System is to improve the core of airborne early warning radar service behaviour.Therefore, before the target that locating and tracking detects, it is necessary first to suppress miscellaneous Clutter present in ripple background or interference.If when under airborne early warning radar depending on being radiated at relatively flat low scattering region, production Raw clutter or interference can be very weak, is handled using conventional method.But generated due to the motion of airborne early warning radar Clutter spectrum spreads in the main lobe broadening and sidelobe clutter of Doppler frequency domain so that shows very strong space-time coupled characteristic, therefore Need to suppress caused clutter or interference using space-time adaptive signal transacting (STAP) technology.Space-time adaptive processing (STAP) technology can make full use of spatial information (si) and time-domain information, and can effective clutter reduction, but as a rule Enough independent same distributions (independent and identically distributed, IID) instruction can not almost be obtained Practice sample to estimate space-time covariance matrix.Even if enough independent same distribution number of training are obtained, for high level matrix There is also the difficulty that amount of calculation and precision aspect are difficult to for the computing inverted.

In the eighties, German doctor R.Klemm is opened up to space-time adaptive signal transacting (STAP) technology Property theoretical research, he, which passes through, carries out thoroughgoing and painstaking analysis to noise performance, finds the big characteristic value of space-time covariance matrix Number be no more than N+M-1, wherein N is the array number of airborne early warning radar, and M is airborne early warning radar in a Coherent processing The umber of pulse of transmitting, illustrates that the full space-time adaptive signal transacting (STAP) for clutter reduction exists at dimensionality reduction really in interval The possibility of reason, on this basis, he proposes accessory channel method (Auxiliary Channel Receiver-ACR), at dimensionality reduction Dimension after reason is by NM to N+M-1.Research show this kind of dimension-reduction treatment in performance close to optimal full space time processing effect, but There is also following two problems in actual applications:First, dimension-reduction treatment is in performance close to optimal full space time processing effect Obtained in the case of without amplitude phase error, if it is considered that spatial domain error, clutter spectrum caused by airborne early warning radar can be along sky Domain Directional Extension so that its clutter dimension is significantly increased, and process performance is decreased obviously;Second, in the array element of airborne early warning radar When number N is bigger, required processor dimension is also bigger.Therefore, accessory channel method still need in actual applications into One-step optimization or improvement, so that clutter recognition performance is improved.

The content of the invention

The problem of existing for above prior art, it is a kind of based on the airborne of accessory channel it is an object of the invention to propose Early warning radar clutter suppression method, this method can solve the problem that tolerance difference and array number of traditional accessory channel method to spatial domain error The problem of being difficult to application when more, and the utilization rate by improving clutter ridge carries out dimension-reduction treatment, so as to reduce required independence With distribution number of training, while recess can be also formed at clutter ridge, improve clutter recognition effect.

To reach above-mentioned technical purpose, the present invention, which adopts the following technical scheme that, to be achieved.

A kind of airborne early warning radar clutter suppression method based on accessory channel, it is characterised in that comprise the following steps:

Step 1, the three-dimensional echo data X that airborne early warning radar receives is obtainedN×M×L, and airborne early warning thunder is obtained accordingly Up to the three-dimensional echo data X receivedN×M×LThe clutter ridge of formation and the main beam spatial domain frequency θ of airborne early warning radar emissions, Then three-dimensional echo data X airborne early warning radar receivedN×M×LRearranged in the way of row, obtain airborne early warning thunder Up to the two-dimentional echo data X receivedNM×L;Wherein, N is the element number of array of airborne early warning radar, and M is airborne early warning radar one The umber of pulse of transmitting in individual coherent processing inteval, L are the three-dimensional echo data X that airborne early warning radar receivesN×M×LRange gate Number;

Step 2, the three-dimensional echo data X received according to airborne early warning radarN×M×LThe clutter ridge of formation, is obtained respectively Num accessory channel and num search passage, and then respectively obtain by each self-corresponding spatial domain frequency structure of num accessory channel Into spatial domain frequency vector θ, the temporal frequency vector that is made up of each self-corresponding temporal frequency of num accessory channelAnd Each self-corresponding time domain steering vector of num search passage, is then calculated dimensionality reduction matrix corresponding to num accessory channel Tb

Step 3, according to the main beam spatial domain frequency θ of airborne early warning radar emissions, airborne early warning radar emission is calculated Main beam spatial domain steering vector Gs, further according to each self-corresponding time domain steering vector of num search passage, appoint to take k-th and search Suo Tongdao time domain steering vector Fk, then according to the main beam spatial domain steering vector G of airborne early warning radar emissionsSearched with k-th Suo Tongdao time domain steering vector Fk, search passage column vector S corresponding to k-th of search passage is calculatedk, and then obtain num Search passage matrix S corresponding to individual search passage;Wherein, k ∈ { 1,2 ... num };

Step 4, the dimensionality reduction square corresponding to search passage matrix S and num accessory channel according to corresponding to num search passage Battle array Tb, optimization dimensionality reduction matrix corresponding to num_s × num_t accessory channel is calculatedAnd then it is calculated based on auxiliary The transformation matrix T of passage;Wherein, k ∈ { 1,2 ... num }, num_s are the optimization dimensionality reduction matrixComprising spatial frequency Number, num_t are the optimization dimensionality reduction matrixComprising Doppler frequency number, num_s × num_t<<num;

Step 5, according to the transformation matrix T based on accessory channel, the two-dimentional echo data received to airborne early warning radar XNM×LWith num search passage corresponding to search passage matrix S carry out dimension-reduction treatment respectively, respectively obtain the number of echoes after dimensionality reduction According to XTWith the search passage steering vector S after dimensionality reductionT

Step 6, according to the echo data X after dimensionality reductionTWith the search passage steering vector S after dimensionality reductionT, dimensionality reduction is calculated Filtering weight vector afterwards;

Step 7, according to the filtering weight vector W after dimensionality reductionT, to the echo data X after dimensionality reductionTClutter recognition processing is carried out, is obtained To range Doppler figure.

Compared with prior art, advantages of the present invention and improvement are:

First, the present invention utilizes clutter ridge information, different dimensionality reduction matrixes can be formed to different search passages, and lead to Cross the clutter that part accessory channel having identical Doppler frequency with search passage, being formed along clutter ridge offsets search passage Component, deficiency of the original auxiliary channel algorithm to spatial domain error tolerance difference is effectively improved, improves practical application.

Second, the present invention utilizes the dimensionality reduction transition matrix based on accessory channel method, the echo received to airborne early warning radar While data carry out doppler filtering processing, dimension-reduction treatment is also carried out, effectively improves original auxiliary channel algorithm due to can The problem of clutter covariance matrix estimation is inaccurate caused by independent same distribution number of training deficiency, and airborne early warning The free degree of radar it is big and caused by amount of calculation is excessive and the problem of equipment cost increase, so as to reduce estimate covariance square Sample number required for battle array so that the present invention can will not also reduce clutter recognition performance while number of training deficiency.

Brief description of the drawings

The present invention is described in further detail with reference to the accompanying drawings and detailed description.

Fig. 1 is a kind of schematic flow sheet of airborne early warning radar clutter suppression method based on accessory channel of the present invention;

Fig. 2 (a) be using the range Doppler figure obtained after pulse Doppler (PD) algorithm process,

Fig. 2 (b) is using the range Doppler figure obtained after the processing of original accessory channel method (ACR);

Fig. 3 is the range Doppler figure obtained after being handled using the inventive method;

Fig. 4 (a) is the global two-dimentional response diagram obtained using the inventive method,

Fig. 4 (b) is the partial enlargement two dimension response diagram obtained using the inventive method;

After Fig. 5 is handles using pulse Doppler (PD) algorithm, original accessory channel method (ACR) and the inventive method respectively Obtained clutter dump power comparison diagram.

Embodiment

Reference picture 1, the flow for a kind of airborne early warning radar clutter suppression method based on accessory channel of the present invention are shown It is intended to, airborne early warning radar clutter suppression method of this kind based on accessory channel, comprises the following steps:

Step 1, the three-dimensional echo data X that airborne early warning radar receives is obtainedN×M×L, and airborne early warning thunder is obtained accordingly Up to the three-dimensional echo data X receivedN×M×LThe clutter ridge of formation and the main beam spatial domain frequency θ of airborne early warning radar emissions, Then three-dimensional echo data X airborne early warning radar receivedN×M×LRearranged in the way of row, obtain airborne early warning thunder Up to the two-dimentional echo data X receivedNM×L;Wherein, N is the element number of array of airborne early warning radar, and M is airborne early warning radar one The umber of pulse of transmitting in individual coherent processing inteval, L are the three-dimensional echo data X that airborne early warning radar receivesN×M×LRange gate Number.

Specifically, airborne early warning radar chooses positive side battle array airborne early warning radar, and positive side battle array airborne early warning radar antenna includes N Individual array element, N number of array element receive the three-dimensional echo data X of ground scatter body reflectionN×M×L, as acquisition airborne early warning radar The three-dimensional echo data X receivedN×M×L;Wherein, N is the element number of array of airborne early warning radar, and M is airborne early warning radar one The umber of pulse of transmitting in individual coherent processing inteval, L are the three-dimensional echo data X that airborne early warning radar receivesN×M×LRange gate Number.

Wherein, because airborne early warning radar uses positive side battle array airborne early warning radar, and the day of positive side battle array airborne early warning radar Linear array member axially it is consistent with the heading of carrier aircraft, the main beam direction of positive side battle array airborne early warning radar illumination ground scatter body and The cone angle cosine value that the bay of positive side battle array airborne early warning radar is axially formed, and the Doppler frequency of ground scatter body echo A kind of linear relationship, and the ground scatter body echo is exactly clutter, the clutter the cone angle cosine value with it is described how general Strangle in frequency space, clutter distribution is straight line, using this straight line as clutter ridge, and obtains airborne early warning radar emission Main beam spatial domain frequency θsFor 0, while the three-dimensional echo data X that airborne early warning radar is receivedN×M×LThe weight in the way of row New arrangement, obtains the two-dimentional echo data X that airborne early warning radar receivesNM×L

Step 2, the three-dimensional echo data X received according to airborne early warning radarN×M×LThe clutter ridge of formation, is obtained respectively Num accessory channel and num search passage, and then respectively obtain by each self-corresponding spatial domain frequency structure of num accessory channel Into spatial domain frequency vector θ, the temporal frequency vector that is made up of each self-corresponding temporal frequency of num accessory channelAnd Each self-corresponding time domain steering vector of num search passage, is then calculated dimensionality reduction matrix corresponding to num accessory channel Tb

Specifically, three-dimensional echo data X airborne early warning radar receivedN×M×LElement number of array and airborne early warning thunder Up to the space for the umber of pulse composition launched in a coherent processing inteval, the three-dimensional echo received as airborne early warning radar Data XN×M×LArray element-pulse domain two-dimensional space, it is and airborne pre- according to the two-dimensional space interior edge of the array element-pulse domain Alert radar receives three-dimensional echo data XN×M×LThe clutter ridge of formation, obtains num accessory channel respectively and num search is logical Road, detailed process are:

If the three-dimensional echo data X that airborne early warning radar receivesN×M×LThe clutter ridge slope of formation is β, airborne early warning thunder Up to the three-dimensional echo data X receivedN×M×LMaximum of the clutter ridge of formation in the frequency of spatial domain is θmax, airborne early warning radar The three-dimensional echo data X receivedN×M×LMinimum value of the clutter ridge of formation in the frequency of spatial domain is θmin, airborne early warning radar connects The three-dimensional echo data X receivedN×M×LMaximum of the clutter ridge of formation in Doppler frequency beAirborne early warning radar The three-dimensional echo data X receivedN×M×LMinimum value of the clutter ridge of formation in Doppler frequency beIts expression formula point It is not:

Wherein, N is the element number of array of airborne early warning radar, and M is that airborne early warning radar is sent out in a coherent processing inteval The umber of pulse penetrated, L are the three-dimensional echo data X that airborne early warning radar receivesN×M×LRange gate number, V be carrier aircraft flight speed Degree, λ are wavelength, frFor pulse recurrence frequency, d represents the adjacent array element interval of airborne early warning radar.

Then, the three-dimensional echo data X received to airborne early warning radarN×M×LThe clutter ridge spatial domain frequency and machine of formation Carry the three-dimensional echo data X that early warning radar receivesN×M×LThe clutter ridge Doppler frequency of formation is evenly dividing respectively, is obtained To num accessory channel and num search passage, and each self-corresponding time domain steering vector of num search passage.

If the spatial domain frequency of i-th of accessory channel is θi, then the temporal frequency of i-th of accessory channel beIts Expression formula is respectively:

The spatial domain frequency vector θ being then made up of each self-corresponding spatial domain frequency of num accessory channel, and it is auxiliary by num The temporal frequency vector for helping each self-corresponding temporal frequency of passage to formExpression formula be respectively:

θ=[θ12,…θnum]

Further according to the spatial domain frequency vector θ being made up of each self-corresponding spatial domain frequency of num accessory channel, and by num The temporal frequency vector that individual each self-corresponding temporal frequency of accessory channel is formedIt is calculated respectively by num accessory channel The spatial domain steering vector column vector G that each self-corresponding spatial domain steering vector is formedbWith by num accessory channel it is each self-corresponding when The time domain steering vector column vector F that domain steering vector is formedb, its expression is respectively:

Gb=[1;ej2πθ;…;ej(N-1)2πθ]

And then dimensionality reduction matrix T corresponding to num accessory channel is calculatedb, its expression formula is:

Wherein, N be airborne early warning radar element number of array, GbTo be oriented to by each self-corresponding spatial domain of num accessory channel The spatial domain steering vector column vector that vector is formed, FbIt is made up of each self-corresponding time domain steering vector of num accessory channel Time domain steering vector column vector, TbFor dimensionality reduction matrix corresponding to num accessory channel,For Kronecker product operation symbol.

Step 3, according to the main beam spatial domain frequency θ of airborne early warning radar emissions, airborne early warning radar emission is calculated Main beam spatial domain steering vector Gs, further according to each self-corresponding time domain steering vector of num search passage, appoint to take k-th and search Suo Tongdao time domain steering vector Fk, then according to the main beam spatial domain steering vector G of airborne early warning radar emissionsSearched with k-th Suo Tongdao time domain steering vector Fk, search passage column vector S corresponding to k-th of search passage is calculatedk, and then obtain num Search passage matrix S corresponding to individual search passage;Wherein, k ∈ { 1,2 ... num }.

Specifically, search passage column vector S corresponding to k-th of search passage in num search passagekIt is according to airborne pre- The main beam spatial domain steering vector G of alert radar emissionsWith the time domain steering vector F of k-th of search passagekIt is calculated, it is expressed Formula is:

Wherein, GsFor the main beam spatial domain steering vector of airborne early warning radar emission, and num search passage is each right The main beam spatial domain steering vector for the airborne early warning radar emission answered is the same;FkIt is oriented to for the time domain of k-th of search passage Vector, SkFor search passage column vector corresponding to k-th of search passage,For Kronecker product operation symbol.

The main beam spatial domain steering vector G of airborne early warning radar emissionsWith the time domain steering vector F of k-th of search passagek Expression formula be respectively:

θs=0

Wherein, θsFor the main beam spatial domain frequency of airborne early warning radar emission,For the time domain frequency of k-th of search passage Rate, N are the element number of array of airborne early warning radar, and M is the umber of pulse that airborne early warning radar is launched in a coherent processing inteval.

According to search passage column vector S corresponding to k-th of search passagek, it is logical to obtain search corresponding to num search passage Road matrix S.

Step 4, the dimensionality reduction square corresponding to search passage matrix S and num accessory channel according to corresponding to num search passage Battle array Tb, optimization dimensionality reduction matrix corresponding to num_s × num_t accessory channel is calculatedAnd then it is calculated based on auxiliary The transformation matrix T of passage;Wherein, k ∈ { 1,2 ... num }, num_s are the optimization dimensionality reduction matrixComprising spatial frequency Number, num_t are the optimization dimensionality reduction matrixComprising Doppler frequency number, num_s ∈ { 1,2 ... num } num_t ∈ 1, 2 ... num }, num_s × num_t<<num.

Specifically, if the Doppler frequency of (i ∈ { 1,2 ... num }) accessory channel is i-th in num accessory channelThen choose num accessory channel in (i- (num_t-1)/2) it is individual to (i+ (num_t-1)/2) individual accessory channel each Corresponding Doppler frequency, as the Doppler frequency of optimization accessory channel, choose (i- (num_s- in num accessory channel 1) it is/2) individual to (i+ (num_s-1)/2) individual each self-corresponding spatial frequency of accessory channel, the space as optimization accessory channel Frequency, then optimize accessory channel and include num_s spatial frequency and num_t Doppler frequency, and obtain optimization auxiliary accordingly The spatial domain frequency vector θ of passages'With the temporal frequency vector of optimization accessory channelIts expression formula is respectively:

θs'=[θi-(num_s-1)/2i-[(num_s-1)/2]+1,…θi,…,θi+[(num_s-1)/2]-1i+(num_s-1)/2]

Typically, num_s, num_t and num relation meet num_s × num_t<<Num, and num_s and num_t difference Take odd number value.

According to the spatial domain frequency vector θ of optimization accessory channels'With the temporal frequency vector of optimization accessory channelRespectively The spatial domain steering vector G of optimization accessory channel is calculatedb'With the time domain steering vector F of optimization accessory channelb', its expression formula Respectively:

And then optimization dimensionality reduction matrix corresponding to num_s × num_t accessory channel is calculatedIts expression formula is:

Wherein, N is the element number of array of airborne early warning radar, and M is that airborne early warning radar is sent out in a coherent processing inteval The umber of pulse penetrated.

Therefore, the transformation matrix T based on accessory channel be according to corresponding to num search passage search passage matrix S and Optimization dimensionality reduction matrix corresponding to num_s*num_t accessory channelIt is calculated, its expression formula is:

Wherein, S is search passage matrix corresponding to num search passage,For num_s × num_t accessory channel pair The optimization dimensionality reduction matrix answered, it is dimensionality reduction matrix T corresponding to num accessory channelbSubset, num_s for optimization accessory channel bag The spatial frequency number contained, num_t are the Doppler frequency number that optimization accessory channel includes, and T is the change based on accessory channel Matrix is changed, H represents conjugate transposition.

Step 5, according to the transformation matrix T based on accessory channel, the two-dimentional echo data received to airborne early warning radar XNM×LWith num search passage corresponding to search passage matrix S carry out dimension-reduction treatment respectively, respectively obtain the number of echoes after dimensionality reduction According to XTWith the search passage steering vector S after dimensionality reductionT

Specifically, according to the transformation matrix T based on accessory channel, the two-dimentional echo data received to airborne early warning radar XNM×LWith num search passage corresponding to search passage matrix S carry out dimension-reduction treatment respectively, obtain the echo data X after dimensionality reductionT With the search passage steering vector S after dimensionality reductionT, its expression formula is respectively:

XT=THXNM×L

ST=THSk

Wherein, XTFor the echo data after dimensionality reduction, T is the transformation matrix based on accessory channel, XNM×LFor airborne early warning thunder Up to the two-dimentional echo data received, STThe search passage steering vector after dimensionality reduction is represented, S is corresponding to num search passage Search passage matrix S, H represent conjugate transposition.

By dimension-reduction treatment back echo data XTDimension by NM dimensionality reductions to ((num_s × num_t)+1), so as to realize machine Carry the dimension-reduction treatment for the two-dimentional echo data that early warning radar receives.

Step 6, according to the echo data X after dimensionality reductionTWith the search passage steering vector S after dimensionality reductionT, utilize maximum likelihood The filtering weight vector after dimensionality reduction is calculated in method.

Specifically, according to the echo data X after dimensionality reductionTWith the search passage steering vector S after dimensionality reductionT, utilize maximum likelihood Echo data X after method estimation dimensionality reductionTCovariance matrix, that is, utilize the echo data X after dimensionality reductionTAs independent same distribution Training sample, estimate to obtain the echo data X after dimensionality reduction by maximum likelihood methodTCovariance matrix RT, its expression formula is:

L1=2 × num_s × num_t

Wherein, RTRepresent the echo data X after dimensionality reductionTCovariance matrix, L1 represents the number of echoes of pending search passage According to required independent same distribution number of samples, XTjAfter representing that the echo data after the dimensionality reduction of j-th of call number, j represent dimensionality reduction Echo data XTJ-th of call number, H represent conjugate transposition.

Adaptively weight vector W calculation formula is:

Wherein, RPerFor the original covariance matrix obtained using traditional accessory channel method, SPerTo be led to using tradition auxiliary The initial search passage steering vector that Dow process obtains.

Then, by the echo data X after dimensionality reductionTCovariance matrix RTThe original obtained instead of using traditional accessory channel method Beginning covariance matrix RPer, with the search passage steering vector S after dimensionality reductionTObtained instead of using traditional accessory channel method original Search passage steering vector SPer, and the filtering weight vector W after dimensionality reduction is calculatedT, its expression formula is:

Wherein, WTRepresent the filtering weight vector after dimensionality reduction, RTFor the echo data X after dimensionality reductionTCovariance matrix, STFor Search passage steering vector after dimensionality reduction.

Step 7, according to the filtering weight vector W after dimensionality reductionT, and using space-time adaptive processing method to the echo after dimensionality reduction Data XTClutter recognition processing is carried out, obtains range Doppler figure.

Specifically, k-th of search passage forms k-th of transformation matrix based on accessory channel method, and num search passage is each The self-corresponding transformation matrix based on accessory channel method, the num different transformation matrixs based on accessory channel method are formed, So as to form num different filtering weight vectors, i.e., the filtering weight vector W after dimensionality reductionT.Therefore, according to the filter after dimensionality reduction Ripple weight vector WT, and (STAP) method is handled to the echo data X after dimensionality reduction using space-time adaptiveTClutter recognition processing is carried out, is obtained Echo data Y to after clutter recognition processing, and the echo data Y outputs after clutter recognition is handled, obtain range Doppler Figure.

The expression formula of echo data Y after clutter recognition processing is:

Y=WT HXT

Wherein, WTFor the filtering weight vector after dimensionality reduction, XTFor the echo data after dimensionality reduction, H represents conjugate transposition.

Make further checking explanation to effect of the present invention with reference to emulation experiment.

(1) echo data emulation and experiment condition:

What the emulation experiment of the present invention was carried out under the softwares of MATLAB 7.11, it is airborne pre- in the emulation experiment of the present invention The antenna of alert radar is using the evenly distributed linear array of 128 array element, and adjacent array element spacing and wavelength d/ λ ratio are 0.5, main ripple The antenna normal direction of Shu Zhixiang airborne early warning radars, i.e., it is 0 ° with front normal direction angle, that is, positive side array antenna. In the emulation experiment of the present invention, the three-dimensional echo data that the airborne early warning radar that uses receives is according to Lincoln laboratory The clutter model simulation that J.Ward is proposed produces, and adds white Gaussian noise, and specific simulation parameter is as shown in table 1:

Table 1

Carrier aircraft height 8km Carrier aircraft speed 150m/s Umber of pulse 128 Wavelength 0.1m Pulse recurrence frequency 8000Hz Antenna axial direction and carrier aircraft velocity angle Miscellaneous noise ratio 70dB Range gate number 500 Main beam points to and bay axial angle 90°

(2) emulation content

In order to illustrate to improve the superiority of accessory channel algorithm, Fig. 2 (a) and Fig. 2 (b) give the place of other several algorithms Result is managed, wherein Fig. 2 (a) is using the range Doppler figure obtained after existing pulse Doppler (PD) algorithm process, Fig. 2 (b) To use the range Doppler figure obtained after the processing of original accessory channel method (ACR).

Fig. 2 (a) and Fig. 2 (b) transverse axis represents Doppler's channel position respectively, and the longitudinal axis represents range gate sequence number respectively;Fig. 2 (a) white portion in is to be distributed using the main clutter dump power obtained after existing pulse Doppler (PD) algorithm process, face The shallower subregion of color is to be distributed using the sidelobe clutter dump power obtained after existing pulse Doppler (PD) algorithm process, black Color region is to be distributed using the noise dump power obtained after existing pulse Doppler (PD) algorithm process;White in Fig. 2 (b) Region is to be distributed using the clutter dump power after the processing of original accessory channel method (ACR), and black region is at existing algorithm ACR Noise dump power distribution after reason.

From Fig. 2 (a) it can be seen that stronger residual spur occupies more distance-Doppler unit area, direct shadow The Radar Targets'Detection in the distance-doppler cells region is rung, from Fig. 2 (b) it can be seen that white portion and color are shallower Region has greatly reduced, and shows that the clutter component in echo data is suppressed.

Fig. 3 is the range Doppler figure obtained after being handled using the inventive method.White portion and color in Fig. 3 is shallower Region has greatly reduced, and shows that the clutter component in the three-dimensional echo data that echo data airborne early warning radar receives obtains Preferably suppress, and independent same distribution number of training is as (2* (num+1) -3) required for original accessory channel method (ACR) It is individual be reduced to the present invention required for (2* (num_s*num_t+1) -3) it is individual, reduce required independent same distribution training sample Number.

Fig. 4 (a) is the global two-dimentional response diagram obtained using the inventive method, and Fig. 4 (b) is to be obtained using the inventive method Partial enlargement two dimension response diagram;Fig. 4 (a) and Fig. 4 (b) transverse axis represents to normalize Doppler frequency, longitudinal axis difference table respectively Show normalization spatial frequency.It can be seen that can be received using the inventive method in airborne early warning radar three from Fig. 4 (b) Tie up and recess is formed at the clutter ridge that echo data is formed.

Fig. 5 is to be contrasted respectively using the clutter dump power obtained after existing algorithm PD, ACR and the inventive method processing Figure;Fig. 5 transverse axis represents Doppler's passage, and the longitudinal axis represents clutter dump power;Solid line in Fig. 5 represents more using existing pulse The clutter dump power obtained after Pu Le (PD) algorithm process, dotted line are to be obtained using after the processing of original accessory channel method (ACR) Clutter dump power, pecked line is obtained clutter dump power after being handled using the inventive method.

The inventive method is lacked relative to what the main lobe gain of original auxiliary channel algorithm reduced as can see from Figure 5, right Spatial domain error tolerance is good, and can utilize less independent same distribution number of training suppressed sidelobes clutter, and main lobe Clutter area narrows, and obtains more preferable clutter recognition effect.

The simulation result shows that the present invention can obtain preferably miscellaneous in the case where reducing training sample and array element error Ripple inhibition.

In summary, emulation experiment demonstrates the correctness of the present invention, validity and reliability.

Obviously, those skilled in the art can carry out the essence of various changes and modification without departing from the present invention to the present invention God and scope;So, if these modifications and variations of the present invention belong to the scope of the claims in the present invention and its equivalent technologies Within, then the present invention is also intended to comprising including these changes and modification.

Claims (10)

1. a kind of airborne early warning radar clutter suppression method based on accessory channel, it is characterised in that comprise the following steps:
Step 1, the three-dimensional echo data X that airborne early warning radar receives is obtainedN×M×L, and obtain airborne early warning radar accordingly and connect The three-dimensional echo data X receivedN×M×LThe clutter ridge of formation and the main beam spatial domain frequency θ of airborne early warning radar emissions, then The three-dimensional echo data X that airborne early warning radar is receivedN×M×LRearranged in the way of row, obtain airborne early warning radar and connect The two-dimentional echo data X receivedNM×L;Wherein, N is the element number of array of airborne early warning radar, and M is airborne early warning radar in a phase The umber of pulse launched in dry-cure interval, L are the three-dimensional echo data X that airborne early warning radar receivesN×M×LRange gate Number;
Step 2, the three-dimensional echo data X received according to airborne early warning radarN×M×LThe clutter ridge of formation, num are obtained respectively Accessory channel and num search passage, and then respectively obtain the sky being made up of each self-corresponding spatial domain frequency of num accessory channel Domain frequency vector θ, the temporal frequency vector being made up of each self-corresponding temporal frequency of num accessory channelAnd num is searched Each self-corresponding time domain steering vectors of Suo Tongdao, are then calculated dimensionality reduction matrix T corresponding to num accessory channelb
Step 3, according to the main beam spatial domain frequency θ of airborne early warning radar emissions, the master of airborne early warning radar emission is calculated Wave beam spatial domain steering vector Gs, further according to each self-corresponding time domain steering vector of num search passage, appoint and take k-th of search to lead to The time domain steering vector F in roadk, then according to the main beam spatial domain steering vector G of airborne early warning radar emissionsIt is logical with k-th of search The time domain steering vector F in roadk, search passage column vector S corresponding to k-th of search passage is calculatedk, and then obtain num and search Search passage matrix S corresponding to Suo Tongdao;Wherein, k ∈ { 1,2 ... num };
Step 4, the dimensionality reduction matrix T corresponding to search passage matrix S and num accessory channel according to corresponding to num search passageb, Optimization dimensionality reduction matrix corresponding to num_s × num_t accessory channel is calculatedAnd then it is calculated based on accessory channel Transformation matrix T;Wherein, k ∈ { 1,2 ... num }, num_s are the optimization dimensionality reduction matrixComprising spatial frequency number, Num_t is the optimization dimensionality reduction matrixComprising Doppler frequency number, num_s × num_t < < num;
Step 5, according to the transformation matrix T based on accessory channel, the two-dimentional echo data X received to airborne early warning radarNM×L With num search passage corresponding to search passage matrix S carry out dimension-reduction treatment respectively, respectively obtain the echo data X after dimensionality reductionT With the search passage steering vector S after dimensionality reductionT
Step 6, according to the echo data X after dimensionality reductionTWith the search passage steering vector S after dimensionality reductionT, after dimensionality reduction is calculated Filter weight vector;
Step 7, according to the filtering weight vector W after dimensionality reductionT, to the echo data X after dimensionality reductionTCarry out clutter recognition processing, obtain away from From Dopplergram.
A kind of 2. airborne early warning radar clutter suppression method based on accessory channel as claimed in claim 1, it is characterised in that In step 2, the spatial domain frequency vector θ being made up of each self-corresponding spatial domain frequency of num accessory channel, auxiliary by num The temporal frequency vector for helping each self-corresponding temporal frequency of passage to formIts expression formula is respectively:
θ=[θ1, θ2... θnum]
Wherein,θmaxFor The three-dimensional echo data X that airborne early warning radar receivesN×M×LMaximum of the clutter ridge of formation in the frequency of spatial domain, θminFor machine Carry the three-dimensional echo data X that early warning radar receivesN×M×LMinimum value of the clutter ridge of formation in the frequency of spatial domain,For machine Carry the three-dimensional echo data X that early warning radar receivesN×M×LMaximum of the clutter ridge of formation in Doppler frequency,For The three-dimensional echo data X that airborne early warning radar receivesN×M×LMinimum value of the clutter ridge of formation in Doppler frequency, N is machine The element number of array of early warning radar is carried, M is the umber of pulse that airborne early warning radar is launched in a coherent processing inteval, and L is airborne The three-dimensional echo data X that early warning radar receivesN×M×LRange gate number.
A kind of 3. airborne early warning radar clutter suppression method based on accessory channel as claimed in claim 1, it is characterised in that In step 2, it is described to obtain dimensionality reduction matrix T corresponding to num accessory channelb, its process is:
According to the spatial domain frequency vector θ being made up of each self-corresponding spatial domain frequency of num accessory channel, and by num auxiliary The temporal frequency vector that each self-corresponding temporal frequency of passage is formedIt is each right by num accessory channel to be calculated respectively The spatial domain steering vector column vector G that the spatial domain steering vector answered is formedbIt is oriented to by each self-corresponding time domain of num accessory channel The time domain steering vector column vector F that vector is formedb, its expression is respectively:
Gb=[1;ej2πθ;...;ej(N-1)2πθ]
And then dimensionality reduction matrix T corresponding to num accessory channel is calculatedb, its expression formula is:
Wherein, N be airborne early warning radar element number of array, GbFor by each self-corresponding spatial domain steering vector structure of num accessory channel Into spatial domain steering vector column vector, FbTime domain to be made up of each self-corresponding time domain steering vector of num accessory channel is led To vector column vector, TbFor dimensionality reduction matrix corresponding to num accessory channel,For Kronecker product operation symbol.
A kind of 4. airborne early warning radar clutter suppression method based on accessory channel as claimed in claim 1, it is characterised in that In step 3, described to obtain search passage matrix S corresponding to num search passage, its process is:
Search passage column vector S corresponding to k-th of search passage in num search passagekIt is according to airborne early warning radar emission Main beam spatial domain steering vector GsWith the time domain steering vector F of k-th of search passagekIt is calculated, its expression formula is:
Wherein, GsFor the main beam spatial domain steering vector of airborne early warning radar emission, and num search passage is each self-corresponding The main beam spatial domain steering vector of airborne early warning radar emission is the same;FkFor the time domain steering vector of k-th of search passage, SkFor search passage column vector corresponding to k-th of search passage,For Kronecker product operation symbol;
The main beam spatial domain steering vector G of airborne early warning radar emissionsWith the time domain steering vector F of k-th of search passagekTable It is respectively up to formula:
θs=0
Wherein, θsFor the main beam spatial domain frequency of airborne early warning radar emission,For the temporal frequency of k-th of search passage, N is The element number of array of airborne early warning radar, M are the umber of pulse that airborne early warning radar is launched in a coherent processing inteval;
According to search passage column vector S corresponding to k-th of search passagek, obtain search passage square corresponding to num search passage Battle array S.
5. a kind of airborne early warning radar clutter suppression method based on accessory channel as claimed in claim 2, its feature exist In in step 4, described to obtain optimization dimensionality reduction matrix corresponding to num_s × num_t accessory channelIts expression formula is:
Wherein,
θs′=[θi-(num_s-1)/2, θi-[(num_s-1)/2]+1... θi..., θi+[(num_s-1)/2]-1, θi+(num_s-1)/2],
Num_s drops for the optimization Tie up matrixComprising spatial frequency number, num_t for it is described optimization dimensionality reduction matrixComprising Doppler frequency number, i ∈ { 1,2 ... num }, N are the element number of array of airborne early warning radar, and M is that airborne early warning radar is launched in a coherent processing inteval Umber of pulse.
A kind of 6. airborne early warning radar clutter suppression method based on accessory channel as claimed in claim 1, it is characterised in that In step 4, the transformation matrix T based on accessory channel, its expression formula are:
T=[SH Tb* H]H
Wherein, S is search passage matrix, T corresponding to num search passageb*For corresponding to num_s × num_t accessory channel Optimize dimensionality reduction matrix, be dimensionality reduction matrix T corresponding to num accessory channelbSubset, num_s for it is described optimization dimensionality reduction matrix Comprising spatial frequency number, num_t for it is described optimization dimensionality reduction matrixComprising Doppler frequency number, T be based on auxiliary The transformation matrix of passage, H represent conjugate transposition.
7. a kind of airborne early warning radar based on accessory channel as described in claim 1 or claim 5 or claim 6 is miscellaneous Ripple suppressing method, it is characterised in that in step 4, the num_s is the optimization dimensionality reduction matrixComprising spatial frequency Number and the num_t are the optimization dimensionality reduction matrixComprising Doppler frequency number, its relation is:
Num_s × num_t < < num
Also, num_s and num_t take odd number value respectively;Wherein, num is search passage number, and num is also auxiliary channel number.
8. a kind of airborne early warning radar clutter suppression method based on accessory channel as claimed in claim 1, its feature exist In, in steps of 5, the echo data X after the dimensionality reductionTWith the search passage steering vector S after dimensionality reductionT, its expression formula difference For:
XT=THXNM×L
ST=THS
Wherein, XTFor the echo data after dimensionality reduction, T is the transformation matrix based on accessory channel, XNM×LReceived for airborne early warning radar The two-dimentional echo data arrived, STThe search passage steering vector after dimensionality reduction is represented, S is that search is logical corresponding to num search passage Road matrix, H represent conjugate transposition.
A kind of 9. airborne early warning radar clutter suppression method based on accessory channel as claimed in claim 1, it is characterised in that In step 6, described to obtain the filtering weight vector after dimensionality reduction, comprising num different filtering weight vectors, its feature is: K-th of search passage forms k-th of transformation matrix based on accessory channel method, and num search passage is each self-corresponding based on auxiliary The transformation matrix of passage method is helped, forms the num different transformation matrixs based on accessory channel method, so as to form num Different filtering weight vector.
10. a kind of airborne early warning radar clutter suppression method based on accessory channel as claimed in claim 1, its feature exist In in step 7, described to obtain range Doppler figure, its process is:
According to the filtering weight vector W after dimensionality reductionT, and STAP methods are handled to the echo data X after dimensionality reduction using space-time adaptiveTEnter The processing of row clutter recognition, the echo data Y after clutter recognition processing is obtained, and the echo data Y after clutter recognition is handled is defeated Go out, obtain range Doppler figure;Echo data Y after the clutter recognition processing, its expression formula are:
Y=WT HXT
Wherein, WTFor the filtering weight vector after dimensionality reduction, XTFor the echo data after dimensionality reduction, H represents conjugate transposition.
CN201510717941.1A 2015-10-29 2015-10-29 Airborne early warning radar clutter suppression method based on accessory channel CN105223557B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201510717941.1A CN105223557B (en) 2015-10-29 2015-10-29 Airborne early warning radar clutter suppression method based on accessory channel

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201510717941.1A CN105223557B (en) 2015-10-29 2015-10-29 Airborne early warning radar clutter suppression method based on accessory channel

Publications (2)

Publication Number Publication Date
CN105223557A CN105223557A (en) 2016-01-06
CN105223557B true CN105223557B (en) 2017-11-21

Family

ID=54992615

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201510717941.1A CN105223557B (en) 2015-10-29 2015-10-29 Airborne early warning radar clutter suppression method based on accessory channel

Country Status (1)

Country Link
CN (1) CN105223557B (en)

Families Citing this family (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105929371B (en) * 2016-04-22 2018-05-04 西安电子科技大学 A kind of airborne radar clutter suppression method based on covariance matrix
CN106772303B (en) * 2016-12-22 2019-02-01 西安电子工程研究所 The channel level clutter suppression method of MTD radar

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
DE3639500A1 (en) * 1986-11-20 1988-06-01 Forschungsgesellschaft Fuer An Radar receiver for mobile radar apparatuses having an antenna array with clutter suppression which acts in a two-dimensional manner
US6297772B1 (en) * 1974-09-23 2001-10-02 The United States Of America As Represented By The Secretary Of The Navy Predicting coherent sidelobe canceller
CN101226236A (en) * 2007-12-18 2008-07-23 西安电子科技大学 Sky wave over-the-horizon radar self-adaption interference rejection method based on sidelobe constraint
CN101561497A (en) * 2009-05-22 2009-10-21 西安电子科技大学 Airborne radar clutter suppression method
CN103954942A (en) * 2014-04-25 2014-07-30 西安电子科技大学 Method for partial combination clutter suppression in airborne MIMO radar three-dimensional beam space
CN104360325A (en) * 2014-11-26 2015-02-18 西安电子科技大学 Space-time adaptive processing method for airborne forward-looking array radar
CN104459658A (en) * 2014-06-19 2015-03-25 西安电子科技大学 Unite-domain positioning space-time two-dimensional self-adaptive processing method based on data fitting

Family Cites Families (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP5193455B2 (en) * 2006-11-06 2013-05-08 株式会社東芝 Radar signal processing device
JP2008157679A (en) * 2006-12-21 2008-07-10 Toshiba Corp Radar signal processor

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US6297772B1 (en) * 1974-09-23 2001-10-02 The United States Of America As Represented By The Secretary Of The Navy Predicting coherent sidelobe canceller
DE3639500A1 (en) * 1986-11-20 1988-06-01 Forschungsgesellschaft Fuer An Radar receiver for mobile radar apparatuses having an antenna array with clutter suppression which acts in a two-dimensional manner
CN101226236A (en) * 2007-12-18 2008-07-23 西安电子科技大学 Sky wave over-the-horizon radar self-adaption interference rejection method based on sidelobe constraint
CN101561497A (en) * 2009-05-22 2009-10-21 西安电子科技大学 Airborne radar clutter suppression method
CN103954942A (en) * 2014-04-25 2014-07-30 西安电子科技大学 Method for partial combination clutter suppression in airborne MIMO radar three-dimensional beam space
CN104459658A (en) * 2014-06-19 2015-03-25 西安电子科技大学 Unite-domain positioning space-time two-dimensional self-adaptive processing method based on data fitting
CN104360325A (en) * 2014-11-26 2015-02-18 西安电子科技大学 Space-time adaptive processing method for airborne forward-looking array radar

Non-Patent Citations (4)

* Cited by examiner, † Cited by third party
Title
"A modified dimension-reduced space-time adaptive processing method";C Luo等;《IEEE Radar Conference》;20141231;第31卷(第5期);第724-728页 *
"Adaptive airborne MTI: an auxiliary channel approach";R Klemm;《Communications Radar and Signal Processing Iee Proceedings F》;19871231;第134卷(第3期);第269-276页 *
"在非均匀环境下辅助通道法STAP处理的性能改善";王万林等;《西安电子科技大学学报(自然科学版)》;20041031;第31卷(第5期);第761-764页 *
"空时二维处理辅助通道法应用研究";杜瑞;《科技信息》;20111231;第767-768页 *

Also Published As

Publication number Publication date
CN105223557A (en) 2016-01-06

Similar Documents

Publication Publication Date Title
Zhang et al. Imaging of a moving target with rotating parts based on the Hough transform
Melvin et al. Assessment of multichannel airborne radar measurements for analysis and design of space-time processing architectures and algorithms
US7369083B2 (en) Method for performing bistatic radar functions
US7154433B1 (en) Method and device for the detection and track of targets in high clutter
Sun et al. Direct data domain STAP using sparse representation of clutter spectrum
US20070200750A1 (en) Unique space time adaptive system (uss)
Gierull et al. Two-step detector for RADARSAT-2's experimental GMTI mode
He et al. Novel range-spread target detectors in non-Gaussian clutter
CN101561497B (en) Airborne radar clutter suppression method
Folster et al. An automotive radar network based on 77 GHz FMCW sensors
Rohling Some radar topics: waveform design, range CFAR and target recognition
CN103257346A (en) Automotive anti-collision radar multi-target detecting method and system
Smith et al. Robust through-the-wall radar image classification using a target-model alignment procedure
Zhang et al. Geometry-information-aided efficient radial velocity estimation for moving target imaging and location based on radon transform
CN103399303B (en) Airborne radar resists intensive deceiving jamming method and system
EP1946145B1 (en) Pulse doppler coherent method and system for snr enhancement
US8232908B2 (en) Inverse synthetic aperture radar image processing
CN103744068B (en) The moving-target detection formation method of dual pathways Continuous Wave with frequency modulation SAR system
Liang Waveform design and diversity in radar sensor networks: theoretical analysis and application to automatic target recognition
CN102156279A (en) Method for detecting moving target on ground by utilizing bistatic radar based on MIMO (Multiple Input Multiple Output)
Chen et al. Radon-fractional ambiguity function-based detection method of low-observable maneuvering target
CN102721947A (en) Efficient three-dimensional space-time adaptive clutter suppression method for airborne digital array radar
EP2533069A1 (en) Signal processing unit and method
Hyun et al. A pedestrian detection scheme using a coherent phase difference method based on 2D range-Doppler FMCW radar
US9594159B2 (en) 2-D object detection in radar applications

Legal Events

Date Code Title Description
PB01 Publication
C06 Publication
SE01 Entry into force of request for substantive examination
C10 Entry into substantive examination
GR01 Patent grant
GR01 Patent grant