CN1218185A - Method for processing space-time multibeam adaptive signals - Google Patents

Abstract

A 2-D (space and time) multi-beam adaptive signal processing method is based on the characteristics of ground noise received by airborne phased-array radar, that is, the power spectrum of noise is distributed by semi-elliptical back shape along the diagonal line as longer axis in 2-D (space and time) plane. After a certain low side-lobe space wave beam and time wave beam, the residual main noise is main-lobe noise and main Doppler pass band noise. At first Bulter beam generator is used to generate multiple noises, then the Doppler filter is casecaded after main beam and adjacement beams, and finally the multiple adjacent Doppler channels of main beam output and multiple adjacent beams are adaptively processed according to the least output power criteria of linear restriction. Its advantages are low freedom, strong robustness, high performance to low-speed target and high anti-interference power.

Description

Method for processing space-time multibeam adaptive signals

The present invention relates to a kind of method for processing space-time multibeam adaptive signals in the signal Processing field, be specially adapted to adopt the signal processing system of the airborne surveillance radar of phased array antenna.

The space-time two-dimensional Adaptive Signal Processing has become an important research direction in airborne surveillance radar field in recent years, is airborne surveillance radar of new generation realizes detecting target from clutter and jamming pattern gordian technique.Airborne radar is owing to the motion of carrier aircraft, and the echo of the different aspect of ground return is different with respect to the speed of carrier aircraft, thereby causes clutter spectrum broadening greatly.The land clutter that general airborne radar receives is coupled when being sky, and this has just determined clutter to suppress must be the space-time two-dimensional filtering problem.The space-time two-dimensional self-adaptive processing can compensate the platform motion effect of airborne radar effectively, and obtains desirable clutter rejection; Simultaneously, the space-time two-dimensional self-adaptive processing influences the factor of signal Processing for some, as near-field scattering, array element excitation error and discrete Active Jamming etc., also can realize compensating adaptively.Because numerous spatial domain and time domain unit number, make that actual to carry out optimum self-adaptive processing be unpractical, the one, operand and equipment amount can't be born, and the 2nd, actual data covariance matrix of carrying out self-adaptive processing also is difficult to formation.

Therefore, the research of space-time two-dimensional Adaptive Signal Processing mainly is the actual implementation method of seeking near optimum handling property, promptly study the method that dimensionality reduction or contraction are handled, present dimension-reduction treatment method mainly is to handle in array element-pulse domain, wave beam-pulse domain, Doppler-array element territory and wave beam-four big territories such as Doppler territory.And based on the disposal route in array element-pulse domain, wave beam-pulse domain, Doppler-array element territory because its spatial domain or time domain are the total spaces, the degree of freedom in system of these methods is still very big, the dimensionality reduction that is carried out is limited; Dimension-reduction treatment method degree of freedom in system in free from error desirable radar system based on wave beam-Doppler territory is lower, and in the actual system that error (particularly array element level error) arranged, its degree of freedom is still bigger.The size of degree of freedom in system directly influences the implementation complexity of system, and particularly for non-homogeneous clutter environment and non-positive side battle array radar, the size of degree of freedom in system directly influences data sampling length and determined the realizability of system.

The objective of the invention is to avoid the weak point in the above-mentioned background technology and a kind ofly take dimensionality reduction respectively in wave beam territory and Doppler territory, and Combined Treatment be suitable for space-time two-dimensional adaptive processing method non-homogeneous clutter environment, the low degree-of-freedom system.And the present invention has that robustness is strong, slower-velocity target detects that performance is good, Project Realization convenient, can take into account characteristics such as anti-Active Jamming simultaneously.

The object of the present invention is achieved like this, and method of the present invention comprises following technical measures:

(1) the rectangle plane phased array antenna of the separable weighting of employing, each row is made microwave during reception

Synthetic, become an equidistant linear array;

(2) the clutter data of spatial domain and time domain are carried out dimensionality reduction, reduce operand;

(3) according to linear restriction minimum output power criterion, the clutter data are carried out the self-adaptation place

Reason;

It is characterized in that also comprising following technical measures step:

(1) use the Bulter Beam-former to reduce the spatial domain dimension;

(2) use discrete Fourier transformation to reduce the time domain dimension;

(3) the optimum self-adaptation power of self-adaptive processing is calculated by following formula:

W(j)=(T ^H _LR _XT _L) ^-1T ^H _LR _XS(ω _si)

Purpose of the present invention can also reach by following technical measures:

(1) use the Bulter Beam-former to reduce the technical measures step of spatial domain dimension: to use wave beam

Form device and form a plurality of wave beams that are less than the aerial array number, promptly main beam and with its mutually

Adjacent wave beam.

(2) use discrete Fourier transformation to reduce the technical measures step of time domain dimension: with discrete Fu

In leaf transformation form and to be lower than a plurality of Doppler's passages that actual pulse is adopted magnetic number, promptly detect

Passage and the Doppler's passage that is adjacent.

(3) the clutter data are carried out the technical measures step of self-adaptive processing: many to main beam only

The clutter of a plurality of adjacent beams outputs of individual adjacent Doppler's passage and sense channel constitutes

Covariance matrix obtains the adaptive weight of respective beam and passage, carries out self-adaptation

Handle.

The present invention compares the following advantage of background technology:

(1) the present invention's adjacent a plurality of wave beams of making up spatial domain and time domain simultaneously realize that jointly two dimension is assorted

Ripple offsets, and essence belongs to the popularization that adjacent multi-beam offsets Active Jamming, can resist simultaneously

Active Jamming is fit to the electronic warfare environment.

(2) the present invention utilizes the phased array radar battle array generally to form multi-beam work and pulse Doppler system

Detect the characteristics of target by Doppler's passage, realizations in directly utilization have now wave beam and

Doppler's passage, and degree of freedom in system is low, so implementation complexity is low, Project Realization

Simply, realizability is strong.

(3) the present invention does not rely on the form of laying of aerial array, is particularly suitable for non-homogeneous clutter ring

The border, therefore robustness of the present invention is good.

(4) the inventive method has application value.

Below in conjunction with drawings and Examples the present invention is described in further detail.

Fig. 1 is an embodiments of the invention electricity functional-block diagram, and Fig. 2 is the structured flowchart of embodiments of the invention.

With reference to Fig. 1, embodiments of the invention are synthetic 1 by aerial array and row submatrix wave beam, mould/number conversion 2, beam-controller 3, digital beam form device 4, discrete Fourier transformation 5, adaptive weighted 6, self-adaptive processing 7 are formed.Wherein embodiment row submatrix wave beam synthetic 1 is made of 16 and the synthetic 1-1 to 1-16 of row submatrix wave beam that connects, 1 mould/number conversion 2-1 to 2-16 of synthetic 1-1 to the 1-16 serial connection of each row submatrix wave beam, discrete Fourier transformation 5 is made of 10 and the discrete Fourier transformation 5-1 to 5-10 that connects, wherein discrete Fourier transformation 5-1 is connected in series 5 adaptive weighted 6-1 to 6-5,1 adaptive weighted 6-6 to 6-14 of each serial connection of all the other discrete Fourier transformation 5-2 to 5-10.The horizontal ordinate of Fig. 2 is a time domain, and ordinate is the spatial domain, and the I district is the main-lobe clutter district, and the II district is Doppler's passband clutter district.

The principle of implementing the inventive method is as follows:

Space-time adaptive is handled and is most importantly tried to achieve adaptive weight, needs to calculate clutter covariance matrix R.The covariance matrix of real system is estimated to obtain, and promptly elder generation records L 2-D data vector sample V in the adjacent unit of detection system _i(i=1,2 ..., L), be calculated as follows the estimated value of R again $\hat{R}:\hat{R}=\frac{1}{L}\underset{i=1}{\overset{L}{\mathrm{\Σ}}}{V}_i{V_i}^H$ 。Can get adaptive weight then $W=\mathrm{\μ}{\hat{R}}^{-1}S,$ Wherein μ is a constant, and S is a steering vector.The selection of contiguous reference unit must be satisfied independent identically distributed condition, simultaneously, is limited in the 3dB in order to make the performance loss that causes by not satisfying condition, and requires L to get 2～3 times degree of freedom in system.If degree of freedom in system is obtained too big, the distance range of sampled data is too big with causing, to such an extent as to homogeneous Gaussian Clutter can seriously be run counter to its default.So necessary strict control system degree of freedom is particularly in the environment of non-homogeneous clutter.This requirement comes down to very harsh for the airborne radar environment.The best way is to adopt the dimension-reduction treatment method.The sampling request of dimension-reduction treatment method is two to three times of dimensionality reduction degree of freedom in system, thereby when selecting less lower dimensional space to handle, its sampling request is lower, is well suited for serious non-homogeneous clutter environment.Independent identically distributed hypothesis has seriously been destroyed in bad border of unconventional electronic interferences, if the distance range that relies on is bigger, its adaptive performance will have bigger loss.

When dimension-reduction treatment, use Bulter Beam-former 4 that the spatial domain is transformed to the wave beam territory from the array element territory, its clutter degree of freedom reduces to numbers of beams p by array number N.Discrete Fourier transformation 5 changes to the Doppler territory with time domain from pulse domain, and its clutter degree of freedom is counted K by impulse sampling and reduced to Doppler's port number q.N and K represent the row submatrix array number in spatial domain and the impulse sampling number of time domain respectively, and N is 16 among the embodiment, and K is 32.P and q represent respectively corresponding to the secondary beam number of sense channel and the adjacency channel number of main channel.P elects 9 as among the embodiment, q elects 5 as, discrete Fourier transformation 5-2 to 5-10 is the discrete Fourier transformation of 9 secondary beam, discrete Fourier transformation 5-1 is the discrete Fourier transformation of main beam, its output is except that sense channel is directly delivered to self-adaptive processing 7, and 5 the Doppler passages adjacent with sense channel are delivered to adaptive weighted 6-1 to 6-5 respectively.Adaptive weighted 6-6 to 6-14 is delivered in the output of corresponding sense channel among the discrete Fourier transformation 5-2 to 5-10.

Conversion through spatial domain and time domain, clutter by the total space transformation of space-time two-dimensional to the subspace, its degree of freedom transforms to p * q dimension from N * K dimension, but still can't satisfy the requirement of real-time processing, thereby only constitute covariance matrix by the clutter of a plurality of adjacent beams outputs of a plurality of adjacent Doppler's passage of main beam and sense channel, and corresponding output is carried out adaptive weighted 6 according to linear restriction minimum output power criterion, this moment, the degree of freedom of system was the p+q dimension, adaptive weighted 6-1 to 6-5 is the weighting of 5 adjacent Doppler's passage outputs of main beam among the embodiment, and adaptive weighted 6-6 to 6-14 is the weighting of 9 adjacent beams outputs of sense channel.After using certain low secondary lobe spatial domain wave beam and time domain " wave beam ", main residual spur is main-lobe clutter district (I district) and Doppler's passband clutter district (II district).The clutter in I district can effectively offset by q adjacent beams of time domain, and the II district can realize disappearing mutually processing by p the adjacent beams in spatial domain.Because the time domain error is very little, generally gets several " wave beams " and gets final product q=5 among the embodiment.And the error effect in spatial domain is bigger, and general II district clutter is expansion to some extent longitudinally, need offset with more wave beam.

The optimum self-adaptation power of adaptive processor of the present invention is calculated by following formula

W (j)=(T ^H _LR _XT _L) ^-1T ^H _LR _XS (ω _Si) (1) R wherein _XBe clutter covariance matrix, T _LBe dimensionality reduction transformation matrix, R _X=E[V (X) V (X) ^H], use during actual computation

Replace.T _LCan be expressed as $T_L={\left[\begin{array}{c}[S_S\left({\mathrm{\ω}}_{\mathrm{Si}}\right)\⊗S_T\left({\mathrm{\ω}}_{t1}\right){]}^T\\ \underset{.}{\overset{.}{.}}\\ [S_S\left({\mathrm{\ω}}_{\mathrm{Si}}\right)\⊗S_T\left({\mathrm{\ω}}_{\mathrm{tq}}\right){]}^T\\ [S_S\left({\mathrm{\ω}}_{S1}\right)\⊗S_T\left({\mathrm{\ω}}_{\mathrm{tj}}\right){]}^T\\ \underset{.}{\overset{.}{.}}\\ [S_S\left({\mathrm{\ω}}_{\mathrm{SP}}\right)\⊗S_T\left({\mathrm{\ω}}_{\mathrm{tj}}\right){]}^T\end{array}\right]}^T-----\left(2\right)$ Wherein, ω _Si(i=1,2 ..., p) and ω _Tj(j=1,2 ..., q) representing normalized radian frequency respectively corresponding to i spatial domain wave beam and j time domain " wave beam ",  is that Kronecher is long-pending, p and q represent respectively corresponding to the secondary beam number of sense channel and the adjacency channel number of main channel.S _s(ω _Si) be the spatial domain steering vector, S _T(ω _Tj) be the time domain steering vector, be respectively $S_T\left({\mathrm{\ω}}_{\mathrm{Si}}\right)=[w_1,w_2{e}^{j{\mathrm{\ω}}_{\mathrm{Si}}},\·\·\·,w_N{e}^{j(N-1){\mathrm{\ω}}_{\mathrm{si}}}{]}^T------\left(3\right)$ $S_T\left({\mathrm{\ω}}_{\mathrm{tj}}\right)={[h_1,h_2{e}^{j{\mathrm{\ω}}_{\mathrm{tj}}},\·\·\·,h_k{e}^{j(k-1){\mathrm{\ω}}_{\mathrm{tj}}}]}^T---\left(4\right)$ { w wherein _n, n=1,2 ..., N} and { h _k, k=1,2 ..., K} is respectively spatial domain and the static weighting coefficient of time domain.

Synthetic 16 row submatrixs of 1 usefulness the wave beam of aerial array and row submatrix wave beam synthesizes the equivalent linear array that planar phased array is synthesized Unit 1 and receives clutter and the signal that ground and target reflection are returned in the embodiment of the invention, deliver to digital beam through mould/number conversion 2 and form device 4, digital beam formation device 4 is transformed into 1 main beam and 9 adjacent secondary beam with the noise signal of 16 unit, deliver to discrete Fourier transformation 5, form main channel and 5 adjacent Doppler's passages through discrete Fourier transformation 5, clutter by a plurality of adjacent beams outputs of a plurality of adjacent Doppler's passage of main beam and sense channel constitutes covariance matrix, try to achieve adaptive weight according to linear restriction minimum output power criterion, and corresponding output is carried out adaptive weighted 6.

The synthetic 1 face battle array with phased array of aerial array and row submatrix wave beam synthesizes linear array and receives, and reduces system's clutter degree of freedom.The effect of beam-controller 3 is the orientation generation wave beam control signals according to echo signal, and control figure Beam-former 4 produces main beam and the adjacent secondary beam corresponding with the echo signal orientation.The effect of mould/number conversion 2 is to be to form device 4 for digital beam after the digital signal to produce multi-beams with the analog signal conversion that the antenna array submatrix is sent here.

Digital beam forms device 4 and 5 pairs of spatial domains of discrete Fourier transformation and time-domain signal and carries out conversion, forms spatial domain multi-beam and time domain hyperchannel, reduces and handles dimension.The clutter of a plurality of adjacent Doppler's passage of 7 pairs of main beam outputs of self-adaptive processing and a plurality of adjacent beams outputs of sense channel constitutes covariance matrix and carries out self-adaptive processing, asks for adaptive weight, and the output after adaptive weighted 6 is sued for peace.

Claims (4)

1. method for processing space-time multibeam adaptive signals comprises following technical measures:

(1) the rectangle plane phased array antenna of the separable weighting of employing, each row is made microwave during reception

Synthetic, become an equidistant linear array;

(2) the clutter data of spatial domain and time domain are carried out dimensionality reduction, reduce operand;

(3) according to linear restriction minimum output power criterion, the clutter data are carried out the self-adaptation place

Reason;

It is characterized in that also comprising following technical measures step:

(1) use the Bulter Beam-former to reduce the spatial domain dimension;

(2) use discrete Fourier transformation to reduce the time domain dimension;

(3) the optimum self-adaptation power of self-adaptive processing is calculated by following formula:

W(j)=(T ^H _LR _XT _L) ^-1T ^H _LR _XS(ω _si)

2. the method for processing space-time multibeam adaptive signals based on phased array antenna according to claim 1, it is characterized in that using the Bulter Beam-former to reduce the technical measures step of spatial domain dimension: to form a plurality of wave beams that are less than the aerial array number with Beam-former, i.e. main beam and the wave beam that is adjacent.

3. the method for processing space-time multibeam adaptive signals based on phased array antenna according to claim 1 and 2, it is characterized in that using discrete Fourier transformation to reduce the technical measures step of time domain dimension: to form a plurality of Doppler's passages that are lower than the actual pulse hits with discrete Fourier transformation, i.e. sense channel and Doppler's passage of being adjacent.

4. according to claim 1 or 2 or 3 described method for processing space-time multibeam adaptive signals based on phased array antenna, it is characterized in that the clutter data are carried out the technical measures step of self-adaptive processing: only the clutter to a plurality of adjacent beams outputs of a plurality of adjacent Doppler's passage of main beam and sense channel constitutes covariance matrix, obtain the adaptive weight of respective beam and passage, carry out self-adaptive processing.

Images