CN102664666A - Efficient robust self-adapting beam forming method of broadband - Google Patents

Abstract

The invention provides an efficient robust self-adapting beam forming method of broadband. The method is applied to the field of wireless communication and comprises steps as follows: performing fast fourier transform (FFT) to received data of an array to obtain the received data on different frequency points and a covariance matrix of the received data of each frequency point; choosing a central frequency point as a reference frequency point; using a propagator thought to respectively performing matrix partitioning on the covariance matrix of each frequency point and the covariance matrix of the central frequency point so as to obtain a propagator of each frequency point and the propagator of the central frequency point; constructing a focusing transformation matrix, focusing the propagators of different frequency points onto the same reference frequency point to obtain the final propagator estimation and noise subspace; and combining with a feature space method to configure a broadband beam forming algorithm weight vector to realize robust self-adapting beam forming of the broadband. In comparison with a traditional coherent signal subspace method, the method does not need any singular value or feature value decomposition, does not need a diagonal loading technique, and can reflect a good performance with respect to an environment having low snapshots and strong desired signals. Particularly, the method has stronger robustness and reduces the complexity under a condition that the desired signal estimation has a certain error.

Description

A kind of broadband efficiently sane adaptive beam formation method

Technical field

The present invention relates to radar, sonar and wireless communication technology field, be specifically related to a kind of broadband efficiently sane adaptive beam formation method.

Background technology

Digital beam forms (DBF) and utilizes Digital Signal Processing, the signal that is received by aerial array is carried out weighting, thereby make signal be able to effective reception through the weights of adjusting each array element.This is that weight vector has directly determined the directional diagram of adaptive array, has promptly determined the reception to useful signal because the weights of each array element are formed the array weight vector.Effective reception to useful signal comprises two aspects: the one, and make array pattern main lobe (the maximum direction of the gain of array antenna) aim at the desired signal direction; The 2nd, make aim at interference signal the zero point of directional diagram effectively to suppress.Traditional beamforming algorithm based on the arrowband when being used for broadband signal, can cause the sensing of wave beam, main lobe width on different frequent points, to change a lot, and causes deviation.Therefore, under a large amount of situation about existing of wideband signal source, how to carry out broadband beams and form, become a research focus in the radio communication.

The broadband DBF method that getting up early occurs is based on incoherent signal subspace method (ISM); It is decomposed into some subbands with broadband signal; On each subband, directly carrying out the arrowband handles; Promptly the signal correlation matrix to each subband carries out narrow-band beam formation, and the weighted value of all subbands is carried out arithmetic average or geometric average, draws broadband signal DBF at last.Usually this type Processing Algorithm can not obtain satisfied result, and main cause is that amount of calculation is big, is unable to estimate the coherent signal source.

(Coherent Signal Subspace Method CSM) is a kind of efficient algorithm to coherent signal-subspace method, at first proposes (Wang by Wang and Kaveh; M Kaveh. " Coherent signal-subspace processing for the detecting and estimation of angles of arrival of multiple wideband sources; " IEEE Transactions on ASSP, vol.33, no.4; 1985, pp.823-831.).This type basic idea is to focus on signal space on nonoverlapping Frequency point in the frequency band on the reference frequency point; Obtain the data covariance matrix of single frequency after the focusing; Utilize the arrowband technology to carry out DBF then, this algorithm can solve the coherent signal source problem.After the CSM algorithm, the CSM class algorithm under many different constraint criterions has been proposed, like TCT (two-sided correlation matrices transformation) algorithm, SST (signal subspace conversion) algorithm, RSS (rotating signal subspace) algorithm or the like.More than these algorithms all need use characteristic value or singular value decomposition at structure focussing matrix the time, operand is o (M ³), excessive operand can cause the difficulty of processing in real time.In addition, these class methods are classical Capton Beam-former or the undistorted response of minimum variance (Minimum Variance Distortionless Response, MVDR) Beam-formers what realize adopting when wave beam forms.This type Beam-former need be known the azimuth information of desired signal in advance; If the arrival bearing to desired signal estimates to be forbidden; The direction that can cause a mistake of main beam pointing; This moment, optimum weights can suppress desired signal as disturbing, and fell into thereby form zero real arrival bearing, and therefore performance is comparatively abominable under the higher environment of signal to noise ratio.The diagonal angle loading technique can improve the performance that wave beam forms to a certain extent, but how accurately choosing diagonal loading amount remains a technical barrier.

Summary of the invention

The present invention is directed to the deficiency of prior art, a kind of broadband sane adaptive beam formation method based on propagation operator and feature space is provided.The present invention combines propagation operator (PM) with the broadband coherent signal-subspace method; Derived based on the broadband signal subspace method of PM; The focussing matrix that this method is constructed does not need singular value decomposition, when obtaining noise subspace, does not need characteristic value decomposition yet, and operand is about o (PM ²).Can find out that the complexity of PM algorithm is about the o of TCT focus method (P/M).Therefore when array number is more, will reduce operand greatly based on the method for propagation operator.At last feature space adaptive beam-forming algorithm thought is applied among the PM, realized that the sane adaptive beam in broadband forms.Compare conventional method, the present invention especially estimates to exist under the condition of certain error at desired signal for embodying preferable performance under the stronger environment of low snap, desired signal, changes method and has stronger robustness.

The present invention realizes that through following technical scheme method step is following:

1) the array received data is carried out the FFT conversion, obtain J the data on the frequency, and then obtain the covariance matrix that each frequency receives data respectively, choose center frequency point f ₀Frequency as a reference;

2) utilize propagation operator thought, respectively with the covariance matrix R of each frequency _jCovariance matrix R with center frequency point ₀Carry out the partitioning of matrix, obtain the propagation operator P of each frequency _jPropagation operator P with center frequency point ₀

3) utilize P _jAnd P ₀Construct the focussing matrix T of each frequency _j, the propagation operator of different frequent points is focused on the same reference frequency point;

4) obtaining final propagation operator estimates

Structure Q and Q ₀, i.e. noise subspace;

5) combine feature space thought, structural wideband beamforming algorithm weight vector w _PM-BESB, realize that sane wideband adaptive wave beam forms.

Below each step of the present invention is done further to specify:

Said step 1), the concrete realization as follows:

Consideration has the situation of the even linear array of M omnidirectional's array element, and array element distance is the half-wavelength of centre frequency.P broadband far-field signal incides on this array with respectively, additional and source signal independent Gaussian white noise.With the data uniform sampling of array received and be divided into the piece of K non-overlapping copies, every comprises N sampled point.Each piece is carried out N point FFT, choose J frequency then and carry out subsequent treatment.Define then the k time snap (k=1 of each discrete frequency

; ..., N) the array received data of j frequency do

x _j，k＝A _j(θ)s _j，k+n _j，k

Here A _j(θ)=[a _j(θ ₁) ..., a _j(θ _P)] expression (the array flow pattern matrix of the dimension of M * P), s _{J, k}Be the signal phasor of (P * 1) dimension, n _{J, k}It is the noise vector of (M * 1) dimension.Choose center frequency point f ₀As focusing on frequency, then the k time snap frequency f ₀The array received data can be written as x _{0, k}=A ₀(θ) s _{0, k}+ n _{0, k}So f _jAnd f ₀The covariance matrix of array received data is on the frequency:

$\left\{\begin{array}{c}{R}_j=\frac{1}{K}{\mathrm{\&Sigma;}}_{k=1}^K{x}_{j,k}{x}_{j,k}^H\\ {\mathrm{<figure-callout\; id="0"\; label="R"\; filenames="HSA00000697442900021.png,HSA00000697442900022.png"\; state="\{\{state\}\}">R</figure-callout>}}_0=\frac{1}{K}{\mathrm{\&Sigma;}}_{k=1}^K{x}_{0,k}{x}_{0,k}^H\end{array}\right.$

Said step 2), the concrete realization as follows:

Respectively with each frequency array covariance matrix R _jAnd center frequency point array covariance matrix R ₀Carrying out piecemeal can get:

R _j＝[G _j，H _j]＝[G _j，G _jP _j]

R ₀＝[G ₀，H ₀]＝[G ₀，G ₀P ₀]

The G here _jAnd G ₀All be (dimension of M * P) matrix, H _jAnd H ₀It all is the dimension of M * (M-P) matrix.Utilize propagation operator thought, covariance matrix carried out the propagation operator that piecemeal obtains each frequency:

$G_j{P}_j=H_j\&DoubleRightArrow;G_j^H{G}_j{P}_j=G_j^H{H}_j$

$\&DoubleRightArrow;P_j={\left(G_j^H{G}_j\right)}^{-1}{G}_j^H{H}_j$

In like manner can draw the propagation operator P of center frequency point ₀

Said step 3), the concrete realization as follows:

Utilize the propagation operator P of each frequency _jPropagation operator P with center frequency point ₀, the definition focussing matrix transforms to the propagation operator of different frequent points on the reference frequency point:

$T_j{P}_j^H=P_0^H$

Then each focusing transform matrix is:

$T_j=P_0^H{P}_j{\left(P_j^H{P}_j\right)}^{-1}$

Said step 4), the concrete realization as follows:

Utilize matrix theory knowledge to be not difficult to obtain following formula:

$H_j{T}_j^H=G_j{P}_j{T}_j^H=G_j{\left(T_j{P}_j^H\right)}^H={\mathrm{<figure-callout\; id="0"\; label="G\; j\; P"\; filenames="HSA00000697442900021.png,HSA00000697442900022.png"\; state="\{\{state\}\}">G</figure-callout>}}_j{P}_{}{}_0$

And then have:

$G_j^H{H}_j{T}_j^H=G_j^H{\mathrm{<figure-callout\; id="0"\; label="G\; j\; P"\; filenames="HSA00000697442900021.png,HSA00000697442900022.png"\; state="\{\{state\}\}">G</figure-callout>}}_j{P}_{}{}_0$

The estimation of the propagation operator of single frequency after finally obtaining focusing on:

${\hat{P}}_0=\frac{1}{J}\left[\underset{j=1}{\overset{J}{\mathrm{\&Sigma;}}}{\left(G_j^H{G}_j\right)}^{-1}{G}_j^H{H}_j{T}_j^H\right]$

In conjunction with propagation operator thought; Utilize the matrix Q of dimension of estimation

the structure M of the propagation operator of single frequency * (M-P), and satisfy:

$Q=\left(\begin{array}{c}{\hat{P}}_0\\ -I_{(M-P)}\end{array}\right)$

And, obtain Q with the Q orthonormalization ₀=Q (Q ^HQ) ^(1/2), be equivalent to the noise subspace of reference frequency point.

Said step 5), the concrete realization as follows:

Utilize the focusing of feature space adaptive beam-forming algorithm thought and propagation operator theoretical, obtain broadband characteristics spatially adaptive wave beam and form (PM-BESB) algorithm weight vector

The R here ₀Refer to the covariance matrix of reference frequency point, a (θ ₀) what represent is the steering vector of desired signal, θ ₀Be the DOA of desired signal.Utilize weight vector to realize that sane wideband adaptive wave beam forms at last.

The present invention has following advantage: 1) compare with traditional coherent signal subspace algorithm, the present invention has reduced computational complexity without any need for singular value or characteristic value decomposition; 2) the present invention does not need the diagonal angle loading technique, has made things convenient for actual treatment; 3) the present invention especially estimates to exist under the condition of certain error at desired signal for embodying preferable performance under the stronger environment of low snap, desired signal, changes method and has stronger robustness.

Description of drawings

Fig. 1 is an algorithm flow structure chart of the present invention

Fig. 2 is the relation of output SINR and the fast umber of beats of frequency domain

Fig. 3-a is the relation (not diagonal angle loading) that the wave beam of TCT-SMI algorithm forms performance and SNR

Fig. 3-b is the relation (loading of 20dB diagonal angle) that the wave beam of TCT-SMI algorithm forms performance and SNR

Fig. 3-c is the relation that the wave beam of PM-BESB algorithm forms performance and SNR

Fig. 3-d is the output SINR of two kinds of algorithms and the relation of SNR

Fig. 4-a is the relation (not diagonal angle loading) that the wave beam of TCT-SMI algorithm forms performance and DOA evaluated error

Fig. 4-b is the relation (loading of 0dB diagonal angle) that the wave beam of TCT-SMI algorithm forms performance and DOA evaluated error

Fig. 4-c is the relation (loading of 20dB diagonal angle) that the wave beam of TCT-SMI algorithm forms performance and DOA evaluated error

Fig. 4-d is the relation that the wave beam of PM-BESB algorithm forms performance and DOA evaluated error

Fig. 4-e is the output SINR of two kinds of algorithms and the relation of DOA evaluated error

Embodiment

The even linear array that instance is formed to 10 omnidirectional's array elements.Three broadband signals, wherein desired signal θ ₀=20 °, two interference signals are respectively from θ ₁=-15 ° and θ ₂=50 °, INR=40dB.The relative bandwidth 40% (1.8GHZ-2.7GHZ) of signal, centre frequency is 2.25GHZ, array element distance is the half-wavelength of centre frequency.The sampled data of accepting is divided into snapshots section (being that the fast umber of beats of frequency domain is snapshots), and every section 64 point (being counting of every section FFT) is chosen 11 frequencies.Contrast is based on the performance of Beam-former with the SMI Beam-former (DL-SMI) of broadband SMI Beam-former that focuses on based on two-sided correlation matrices transformation (TCT) and diagonal angle loading technique of PM-BESB, and the result adopts the mean value of 100 independent experiments.This instance is estimated to exist under three kinds of situation of error through the fast umber of beats of different frequency domains, different desired signal signal to noise ratio and to desired signal DOA, relatively the performance of PM-ESB algorithm and the formation of traditional MVDR wave beam.Concrete implementation procedure is following:

1) the fast umber of beats of frequency domain is to Effect on Performance

Fixing desired signal SNR=20dB has investigated the output Signal to Interference plus Noise Ratio of two kinds of algorithm arrays under fast umber of beats snapshots from 10 to 100 conditions of frequency domain, and is as shown in Figure 1.It is thus clear that the TCT-SMI algorithm that PM-BESB algorithm and 20dB diagonal angle load has output SINR much at one, then performance is relatively poor relatively without TCT-SMI that the diagonal angle loads.

2) desired signal SNR is to Effect on Performance

Fixedly the fast umber of beats of frequency domain is 50, has investigated the performance of two kinds of wave beam formation methods and the relation of desired signal SNR.Fig. 2-a is depicted as the TCT-SMI algorithm does not have the performance under the diagonal angle loading environment, and visible when SNR is low (SNR=0dB), the SMI algorithm has good performance.But along with the rising of SNR, secondary lobe is raised gradually.When SNR=20dB, at main lobe θ ₀=20 ° of directions begin to occur zero and fall into, and explain to have occurred expecting the phenomenon of signal cancellation this moment.Fig. 2-b is depicted as the performance of TCT-SMI algorithm under 20dB diagonal angle loading environment, and it is very stable to find out that wave beam forms.The PM-ESB algorithm is (SNR=0dB) under the low signal-to-noise ratio condition; The wave beam minor level is higher relatively; Main cause is because under the lower situation of signal to noise ratio, and the error of structure propagation operator and focusing transform matrix is bigger, has a comparatively sane performance yet the high s/n ratio condition is next; From Fig. 2-c, can see the variation of PM-BESB algorithm with SNR, its beam pattern is relatively stable.Fig. 2-d is the output SINR of two kinds of methods and the relation of SNR.

3) desired signal DOA error is to Effect on Performance

Fixedly the fast umber of beats of frequency domain is 50, desired signal SNR=10dB.The real DOA of desired signal is θ ₀=20 °, but because the existence of systematic error and noise makes the DOA that estimates between ± 5 °, change.Fig. 3-(a～d) is the relation that the wave number of two kinds of methods forms performance and desired signal DOA evaluated error; Fig. 3-e is the output SINR of two kinds of methods and the relation of desired signal DOA evaluated error; Can find out; Performance without the TCT-SMI algorithm of diagonal angle loading processing descends along with the increase of angle estimated bias fast, and the diagonal angle loads the robustness that can promote the SMI algorithm greatly, but still can't overcome the influence that the desired signal evaluated error is pointed to the wave beam main lobe.The wave beam formation performance and the output SINR that can find out the PM-BESB Beam-former among the result change with the variation of DOA evaluated error hardly.

Tradition is based on the broadband MVDR Beam-former of coherent signal subspace focusing algorithm, and under no diagonal angle loading environment, it is very abominable that wave beam forms performance.The diagonal angle loads and necessarily improves the wave beam performance on the program, but still very responsive for the error of desired signal DOA estimation, is mainly reflected in the skew that the wave beam main lobe points to.It is pointed out that actual desired signal generally all is unknown, this accurate estimation that just causes for the desired signal arrival bearing is very important to the performance that broadband beams forms.In addition, in the reality diagonal loading amount choose often relevantly with signal characteristic and surrounding environment, these characteristics have restricted processing more flexible.The present invention has reduced computational complexity without any need for singular value or characteristic value decomposition; In addition, the present invention loads measure without any need for the diagonal angle, and for all embodying preferable performance under the stronger environment of low snap, desired signal, especially estimates to exist under the condition of certain error at desired signal, changes method and has stronger robustness.

Claims (9)

1. the sane adaptive beam in broadband formation method efficiently is characterized in that may further comprise the steps:

1) the array received data is carried out the FFT conversion, obtain J the data on the frequency, and then obtain the covariance matrix that each frequency receives data respectively, choose center frequency point f ₀Frequency as a reference;

2) utilize propagation operator thought, respectively with the covariance matrix R of each frequency _jCovariance matrix R with center frequency point ₀Carry out the partitioning of matrix, obtain the propagation operator P of each frequency _jPropagation operator P with center frequency point ₀

3) utilize P _jAnd P ₀Construct the focussing matrix T of each frequency _j, the propagation operator of different frequent points is focused on the same reference frequency point;

4) obtaining final propagation operator estimates

Structure Q and Q ₀, i.e. noise subspace;

5) combine feature space thought, structural wideband beamforming algorithm weight vector w _PM-BESB, realize that sane wideband adaptive wave beam forms.

2. a kind of broadband efficiently according to claim 1 sane adaptive beam formation method is characterized in that, said step 1), and concrete the realization as follows:

Consideration has the situation of the even linear array of M omnidirectional's array element, and array element distance is the half-wavelength of centre frequency.P broadband far-field signal incides on this array with

respectively, additional and source signal independent Gaussian white noise.With the data uniform sampling of array received and be divided into the piece of K non-overlapping copies, every comprises N sampled point.Each piece is carried out N point FFT, choose J frequency then and carry out subsequent treatment.Define then the k time snap (k=1 of each discrete frequency ; ..., N) the array received data of j frequency do

x _j，k＝A _j(θ)s _j，k+n _j，k

Here A _j(θ)=[a _j(θ ₁) ..., a _j(θ _P)] expression (the array flow pattern matrix of the dimension of M * P), s _{J, k}Be the signal phasor of (P * 1) dimension, n _{J, k}It is the noise vector of (M * 1) dimension.

3. a kind of broadband efficiently according to claim 2 sane adaptive beam formation method is characterized in that, after obtaining each frequency reception data, chooses center frequency point f ₀As focusing on frequency, then the k time snap frequency f ₀The array received data can be written as x _{0, k}=A ₀(θ) s _{0, k}+ n _{0, k}So f _jAnd f ₀The covariance matrix of array received data is on the frequency:

$\left\{\begin{array}{c}{R}_j=\frac{1}{K}{\mathrm{\&Sigma;}}_{k=1}^K{x}_{j,k}{x}_{j,k}^H\\ R_0=\frac{1}{K}{\mathrm{\&Sigma;}}_{k=1}^K{x}_{0,k}{x}_{0,k}^H\end{array}\right.$

4. a kind of broadband sane adaptive beam formation method efficiently according to claim 1 is characterized in that said step 2), the concrete realization as follows:

Respectively with each frequency array covariance matrix R _jAnd center frequency point array covariance matrix R ₀Carrying out piecemeal can get:

R _j＝[G _j，H _j]＝[G _j，G _jP _j]

R ₀＝[G ₀，H ₀]＝[G ₀，G ₀P ₀]

The G here _jAnd G ₀All be (dimension of M * P) matrix, H _jAnd H ₀It all is the dimension of M * (M-P) matrix.

5. a kind of broadband sane adaptive beam formation method efficiently according to claim 4 is characterized in that, utilizes propagation operator thought, and covariance matrix is carried out the propagation operator that piecemeal obtains each frequency:

$G_j{P}_j=H_j\&DoubleRightArrow;G_j^H{G}_j{P}_j=G_j^H{H}_j$

$\&DoubleRightArrow;P_j={\left(G_j^H{G}_j\right)}^{-1}{G}_j^H{H}_j$

In like manner can draw the propagation operator P of center frequency point ₀

6. a kind of broadband sane adaptive beam formation method efficiently according to claim 1 is characterized in that, said step 3), and concrete the realization as follows:

Utilize the propagation operator P of each frequency _jPropagation operator P with center frequency point ₀, the definition focussing matrix transforms to the propagation operator of different frequent points on the reference frequency point:

$T_j{P}_j^H=P_0^H$

Then each focusing transform matrix is:

$T_j=P_0^H{P}_j{\left(P_j^H{P}_j\right)}^{-1}$

7. a kind of broadband efficiently according to claim 1 sane adaptive beam formation method is characterized in that, said step 4), and concrete the realization as follows:

Utilize matrix theory knowledge to be not difficult to obtain following formula:

$H_j{T}_j^H=G_j{P}_j{T}_j^H=G_j{\left(T_j{P}_j^H\right)}^H=G_j{P}_0$

And then have:

$G_j^H{H}_j{T}_j^H=G_j^H{G}_j{P}_0$

The estimation of the propagation operator of single frequency after finally obtaining focusing on:

${\hat{P}}_0=\frac{1}{J}\left[\underset{j=1}{\overset{J}{\mathrm{\&Sigma;}}}{\left(G_j^H{G}_j\right)}^{-1}{G}_j^H{H}_j{T}_j^H\right]$

8. a kind of broadband efficiently according to claim 7 sane adaptive beam formation method; It is characterized in that; In conjunction with propagation operator thought; Utilize the matrix Q of dimension of estimation

the structure M of the propagation operator of single frequency * (M-P), and satisfy:

$Q=\left(\begin{array}{c}{\hat{P}}_0\\ -I_{(M-P)}\end{array}\right)$

And, obtain Q with the Q orthonormalization ₀=Q (Q ^HQ) ^(1/2), be equivalent to the noise subspace of reference frequency point.

9. a kind of broadband efficiently according to claim 1 sane adaptive beam formation method is characterized in that, said step 5), and concrete the realization as follows:

Utilize the focusing of feature space adaptive beam-forming algorithm thought and propagation operator theoretical, obtain broadband characteristics spatially adaptive wave beam and form (PM-BESB) algorithm weight vector

The R here ₀Refer to the covariance matrix of reference frequency point, a (θ ₀) what represent is the steering vector of desired signal, θ ₀Be the DOA of desired signal.Utilize weight vector to realize that sane wideband adaptive wave beam forms at last.

Images