CN103245934A - Beam forming method under large dynamic desired signal - Google Patents

Abstract

The invention discloses a beam forming method under a large dynamic desired signal. The beam forming method under the large dynamic desired signal comprises the following implementation steps of: (1) representing a signal received by an array at the k<th> moment as x(k); (2) performing orthogonal transformation on a received real signal to obtain an analytic signal; (3) setting an initial value of a weight vector as a direction vector v0 of a reference angle, and performing an iterative computation according to a formula, wherein a converged weight vector is wf(K); (4) calculating a directional diagram according to the obtained the weight vector is wf(K); (5) estimating a differential coefficient which corresponds to the reference angle in the directional diagram; (6) judging: selecting a weight vector wq of a phased array as a beam forming weight vector if the absolute value of alpha is greater than a threshold alpha0, i.e. we=wf(K), and selecting the weight vector wf(K) as the beam forming weight vector if the absolute value of alpha is not greater than a threshold alpha0, i.e. we=wf(K); and (7) calculating the output of beam forming. According to the beam forming method under the large dynamic desired signal disclosed by the invention, the self-null problem is solved. Compared with various complex robust beam forming algorithms, the beam forming method under the large dynamic desired signal is simple to implement and meets the requirement of engineering application.

Description

Beam synthesizing method under a kind of big dynamic desirable signal

Technical field

The present invention relates to the Array Signal Processing field, be specifically related to the beam synthesizing method under a kind of big dynamic desirable signal in the wireless receiver.

Technical background

Wave beam is synthetic to be a kind of airspace filter, and it can be applied in the systems such as radar, sonar, radio communication, and it mainly acts on is the wanted signal that receives specific direction, and weakens the undesired signal of other direction.In real system, (Direction Of Arrival be known DOA), but this angle certainly exists certain error to the direction of arrival of wanted signal.The Frost algorithm is the synthetic traditional rudimentary algorithm of adaptive beam.Be the big dynamically application of input for a wanted signal, in desired signal power hour, this DOA error is to the not influence of Frost algorithm.But if wanted signal self power ratio is bigger, the DOA error causes the Frost algorithm to produce " falling into from zero " problem, expects that namely signal is regarded as undesired signal and is suppressed.At this moment, array pattern produces zero in the wanted signal direction and falls into, and the adaptive beam composition algorithm lost efficacy.Now at this problem, though exist various sane adaptive filter algorithms to solve, such as the diagonal angle loading algorithm, the derivative constraints algorithm, convex set is optimized algorithm etc., and these algorithm computation processes are quite complicated, can't be used in engineering.At wanted signal DOA error influential characteristics when the large-signal power only, adopt directional diagram is detected judgement, if produce " falling into from zero ", then adopt the phased array weighting, thereby can address this problem.

Summary of the invention

Technical matters to be solved by this invention is: when wanted signal has greatly dynamically, because there is error in wanted signal DOA, the wanted signal of high-power input causes Frost adaptive beam composition algorithm to lose efficacy, and existing sane wave beam composition algorithm is because calculation of complex is difficult to realization on the engineering.

The present invention is the beam synthesizing method under a kind of big dynamic desirable signal, and implementation step is:

The first step, to the even linear array that N array element constitutes, array element distance d is that the reception signal of half wavelength λ/2, the k moment array can be expressed as $x\left(k\right)=\left[\begin{array}{c}{x}_1\left(k\right)\\ \&CenterDot;\&CenterDot;\&CenterDot;\\ x_N\left(k\right)\end{array}\right],$ N is natural number, k=1 ...;

In second step, the signal x (k) that k is received constantly carries out orthogonal transformation, and real signal is become corresponding analytic signal $\hat{x}\left(k\right)=\left[\begin{array}{c}{\hat{x}}_1\left(k\right)\\ \&CenterDot;\&CenterDot;\&CenterDot;\\ {\hat{x}}_N\left(k\right)\end{array}\right];$

The 3rd step is according to analytic signal

Reference value with the known desired direction of arrival Press the Frost algorithm and calculate weight vector w _f:

(1) calculates with reference to angle

Direction vector $v_{}$ ${}_0=\left[\begin{array}{c}1\\ e^{\mathrm{j\&pi;}\sin {\widehat{\mathrm{\&theta;}}}_0}\\ \&CenterDot;\&CenterDot;\&CenterDot;\\ e^{\mathrm{j\&pi;}(N-1)\sin {\widehat{\mathrm{\&theta;}}}_0}\end{array}\right],$ And definite step size mu;

(2) calculate N * N dimension matrix P=I-v ₀(v ₀ ^Hv ₀) ^-1v ₀ ^H, I is that N * N ties up unit matrix;

(3) calculate the N dimensional vector

(4) get variable m=1, weight vector initial value w _f(1)=w _q

(5) calculate the m time output Symbol H represents conjugate transpose;

(6) upgrade weight vector $w_f(m+1)=P(w_f\left(m\right)-\mathrm{\&mu;}\hat{x}\left(m\right)y^{*}\left(m\right))+w_q;$

(7) make m=m+1, repeat above-mentioned (5)-(7) step, up to weight vector w _fTill the convergence, this moment m=K;

In the 4th step, calculating at this moment, array pattern is

The 5th step arranged angular deflection constant △ θ, estimated on the directional diagram with reference to angle

Corresponding derivative $\mathrm{\&alpha;}=\frac{B({\widehat{\mathrm{\&theta;}}}_0+\mathrm{\&Delta;\&theta;})-B({\widehat{\mathrm{\&theta;}}}_0-\mathrm{\&Delta;\&theta;})}{2\mathrm{\&Delta;\&theta;}};$

In the 6th step, threshold value α is set ₀, if the absolute value of α is greater than thresholding α ₀, then being illustrated near the directional diagram of desired orientation has zero to fall into, and selects vector w _qAs last wave beam synthetic weight vector, i.e. w _e=w _qOtherwise, the weight vector w that uses the Frost algorithm to obtain _f(K) as last wave beam synthetic weight vector, i.e. w _e=w _f(K);

The 7th step, the synthetic output of array beams

This moment k=K+1, K+2 ....

The invention has the beneficial effects as follows: new method at first adopts traditional F rost adaptive beam composition algorithm to obtain array pattern, judge according to directional diagram whether the wanted signal direction produces zero and fall into, if produce, illustrate that desired signal power is bigger, the synthetic phased array weighting scheme that adopts of wave beam this moment, do not fall into the synthetic weight vector that adopts traditional F rost algorithm to calculate of wave beam if produce zero.Than the sane wave beam composition algorithm of various complexity, the inventive method realizes simple, satisfies the engineering demands of applications.

Description of drawings

Fig. 1 is the synoptic diagram of array under the even linear array;

Fig. 2 is wave beam composite structure block diagram;

Fig. 3 is the process flow diagram of the inventive method;

Fig. 4 is the directional diagram of Frost algorithm under the high-power wanted signal;

Specific implementation method

Beam synthesizing method under a kind of big dynamic desirable signal, this method at first adopt the Frost algorithm of traditional adaptive beam in synthetic to calculate weight vector w _f, then to this moment the wanted signal direction array pattern carry out analysis and judgement.Fall into if directional diagram occurs zero in the wanted signal direction, illustrate that signal power is strong, adopt phased array weight vector w this moment _qAs the synthetic weight vector w of wave beam _eThough, phased array weight vector w _qCan't suppress fully undesired signal, still, by the array side lobe attenuation, array output can reach engine request.Do not fall into if occur zero, illustrate that the Frost algorithm is effectively, the weight vector w that the Frost algorithm generates _fAs the synthetic weight vector w of wave beam _e

Fig. 1 is the synoptic diagram of array under the even linear array.Array is made of N identical array element, and array element distance d is half wavelength λ/2.There is a wanted signal in the space, and (Direction of Arrival DOA) is θ to direction of arrival ₀, J undesired signal arranged simultaneously, direction of arrival DOA is respectively θ _j, j=1,2 ..., J, each DOA is positioned at the interval

J＜N-1.In the even linear array that systems array constitutes for N array element, it is reference array element that first array element is set, and is positioned at true origin, is s when then j signal arrives this _j(t).The k reception signal of array constantly can be expressed as $x\left(k\right)=\left[\begin{array}{c}{x}_1\left(k\right)\\ \&CenterDot;\&CenterDot;\&CenterDot;\\ x_N\left(k\right)\end{array}\right],$ Wherein first array element is received to such an extent that signal can be expressed as n ₁(k) noise signal that receives for first array element, then l array element reception signal is

τ _lThe time-delay of the signal that to be the signal that receives of l array element receive with respect to first array element, n _l(k) be the noise signal that l array element receives.

Fig. 2 is for the synthetic structured flowchart of wave beam, and according to nyquist sampling theorem, the real signal x (t) that aerial array receives obtains digital real signal x (k) through sampling, and real signal x (k) becomes analytic signal through orthogonal transformation

Analytic signal

With weight vector w _e(k) weighted sum obtains array output The weight vector generation module determines that wave beam synthesizes weight vector w _e(k) be to adopt phased array weight vector w _qOr Frost algorithm weight vector w _f

Fig. 3 is the schematic flow sheet of the inventive method, and concrete grammar is divided into seven steps:

The first step, to the even linear array that N array element constitutes, array element distance d is that the signal of half wavelength λ/2, the k moment array received can be expressed as $x\left(k\right)=\left[\begin{array}{c}{x}_1\left(k\right)\\ \&CenterDot;\&CenterDot;\&CenterDot;\\ x_N\left(k\right)\end{array}\right],$ N is natural number, k=1 ...;

In second step, the signal x (k) that k is received constantly carries out orthogonal transformation, and real signal is become corresponding analytic signal $\hat{x}\left(k\right)=\left[\begin{array}{c}{\hat{x}}_1\left(k\right)\\ \&CenterDot;\&CenterDot;\&CenterDot;\\ {\hat{x}}_N\left(k\right)\end{array}\right];$

The 3rd step is according to analytic signal

Reference value with the known desired direction of arrival Press the Frost algorithm and calculate weight vector w _f:

(1) calculates with reference to angle Direction vector $v_{}$ ${}_0=\left[\begin{array}{c}1\\ e^{\mathrm{j\&pi;}\sin {\widehat{\mathrm{\&theta;}}}_0}\\ \&CenterDot;\&CenterDot;\&CenterDot;\\ e^{\mathrm{j\&pi;}(N-1)\sin {\widehat{\mathrm{\&theta;}}}_0}\end{array}\right],$ And definite step size mu;

(2) calculate N * N dimension matrix P=I-v ₀(v ₀ ^Hv ₀) ^-1v ₀ ^H, I is that N * N ties up unit matrix;

(3) calculate the N dimensional vector

(4) get variable m=1, weight vector initial value w _f(1)=w _q

(5) calculate the m time output

Symbol H represents conjugate transpose;

(6) upgrade weight vector $w_f(m+1)=P(w_f\left(m\right)-\mathrm{\&mu;}\hat{x}\left(m\right)y^{*}\left(m\right))+w_q;$

(7) make m=m+1, repeat above-mentioned (5)-(7) step, up to weight vector w _fTill the convergence, this moment m=K;

In the 4th step, calculating at this moment, array pattern is

The 5th step arranged angular deflection constant △ θ, estimated on the directional diagram with reference to angle

Corresponding derivative $\mathrm{\&alpha;}=\frac{B({\widehat{\mathrm{\&theta;}}}_0+\mathrm{\&Delta;\&theta;})-B({\widehat{\mathrm{\&theta;}}}_0-\mathrm{\&Delta;\&theta;})}{2\mathrm{\&Delta;\&theta;}};$

In the 6th step, threshold value α is set ₀, if the absolute value of α is greater than thresholding α ₀, then being illustrated near the directional diagram of desired orientation has zero to fall into, and selects vector w _qAs last wave beam synthetic weight vector, i.e. w _e=w _qOtherwise, the weight vector w that uses the Frost algorithm to obtain _f(K) as last wave beam synthetic weight vector, i.e. w _e=w _f(K);

The 7th step, the synthetic output of array beams

This moment k=K+1, K+2 ....

Fig. 4 is the directional diagram of Frost algorithm under the high-power wanted signal, is used for the implication that explanation " falls into from zero ", and wherein the actual DOA of wanted signal is 35 °, and the known wanted signal of system is 30 ° with reference to DOA.Calculate weight vector according to the Frost algorithm, and obtain directional diagram.

The invention has the beneficial effects as follows: new method at first adopts traditional F rost adaptive beam composition algorithm to obtain array pattern, judge according to directional diagram whether the wanted signal direction produces zero and fall into, if produce, illustrate that desired signal power is bigger, " falling into from zero " phenomenon appears, the synthetic phased array weighting scheme that adopts of wave beam this moment does not fall into if produce zero, the synthetic weight vector that adopts traditional F rost algorithm to calculate of wave beam.The inventive method has not only overcome " falling into from zero " problem, and than the sane wave beam composition algorithm of various complexity, this method realizes simple, satisfies the engineering demands of applications.The present invention can be applied in the wireless receiver in fields such as radar, communication.

Claims (1)

1. the beam synthesizing method under one kind big dynamic desirable signal, implementation step is:

The first step, to the even linear array that N array element constitutes, array element distance d is that the reception signal of half wavelength λ/2, the k moment array can be expressed as $x\left(k\right)=\left[\begin{array}{c}{x}_1\left(k\right)\\ \&CenterDot;\&CenterDot;\&CenterDot;\\ x_N\left(k\right)\end{array}\right],$ N is natural number, k=1 ...;

In second step, the signal x (k) that k is received constantly carries out orthogonal transformation, and real signal is become corresponding analytic signal $\hat{x}\left(k\right)=\left[\begin{array}{c}{\hat{x}}_1\left(k\right)\\ \&CenterDot;\&CenterDot;\&CenterDot;\\ {\hat{x}}_N\left(k\right)\end{array}\right];$

The 3rd step is according to analytic signal Reference value with the known desired direction of arrival

Press the Frost algorithm and calculate weight vector w _f:

(1) calculates with reference to angle Direction vector $v_0=\left[\begin{array}{c}1\\ e^{\mathrm{j\&pi;}\sin {\widehat{\mathrm{\&theta;}}}_0}\\ \&CenterDot;\&CenterDot;\&CenterDot;\\ e^{\mathrm{j\&pi;}(N-1)\sin {\widehat{\mathrm{\&theta;}}}_0}\end{array}\right],$ And definite step size mu;

(2) calculate N * N dimension matrix P=I-v ₀(v ₀ ^Hv ₀) ^-1v ₀ ^H, I is that N * N ties up unit matrix;

(3) calculate the N dimensional vector

(4) get variable m=1, weight vector initial value w _f(1)=w _q

(5) calculate the m time output

Symbol H represents conjugate transpose;

(6) upgrade weight vector $w_f(m+1)=P(w_f\left(m\right)-\mathrm{\&mu;}\hat{x}\left(m\right)y^{*}\left(m\right))+w_q;$

(7) make m=m+1, repeat above-mentioned (5)-(7) step, up to weight vector w _fTill the convergence, this moment m=K;

In the 4th step, calculating at this moment, array pattern is

The 5th step arranged angular deflection constant △ θ, estimated on the directional diagram with reference to angle

Corresponding derivative $\mathrm{\&alpha;}=\frac{B({\widehat{\mathrm{\&theta;}}}_0+\mathrm{\&Delta;\&theta;})-B({\widehat{\mathrm{\&theta;}}}_0-\mathrm{\&Delta;\&theta;})}{2\mathrm{\&Delta;\&theta;}};$

In the 6th step, threshold value α is set ₀, if the absolute value of α is greater than thresholding α ₀, then being illustrated near the directional diagram of desired orientation has zero to fall into, and selects vector w _qAs last wave beam synthetic weight vector, i.e. w _e=w _qOtherwise, the weight vector w that uses the Frost algorithm to obtain _f(K) as last wave beam synthetic weight vector, i.e. w _e=w _f(K);

The 7th step, the synthetic output of array beams

This moment k=K+1, K+2 ....

Images