CN108896954B - Estimation method of angle of arrival based on joint real-value subspace in co-prime matrix - Google Patents

Estimation method of angle of arrival based on joint real-value subspace in co-prime matrix Download PDF

Info

Publication number
CN108896954B
CN108896954B CN201810579211.3A CN201810579211A CN108896954B CN 108896954 B CN108896954 B CN 108896954B CN 201810579211 A CN201810579211 A CN 201810579211A CN 108896954 B CN108896954 B CN 108896954B
Authority
CN
China
Prior art keywords
subarray
sub
matrix
subspace
doa
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Active
Application number
CN201810579211.3A
Other languages
Chinese (zh)
Other versions
CN108896954A (en
Inventor
李建峰
蒋德富
沈明威
王峰
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Hohai University HHU
Original Assignee
Hohai University HHU
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 Hohai University HHU filed Critical Hohai University HHU
Priority to CN201810579211.3A priority Critical patent/CN108896954B/en
Publication of CN108896954A publication Critical patent/CN108896954A/en
Application granted granted Critical
Publication of CN108896954B publication Critical patent/CN108896954B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
    • G01S3/00—Direction-finders for determining the direction from which infrasonic, sonic, ultrasonic, or electromagnetic waves, or particle emission, not having a directional significance, are being received
    • G01S3/02—Direction-finders for determining the direction from which infrasonic, sonic, ultrasonic, or electromagnetic waves, or particle emission, not having a directional significance, are being received using radio waves
    • G01S3/14—Systems for determining direction or deviation from predetermined direction

Landscapes

  • Physics & Mathematics (AREA)
  • Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Radar, Positioning & Navigation (AREA)
  • Remote Sensing (AREA)
  • Radar Systems Or Details Thereof (AREA)

Abstract

The invention discloses a Direction of arrival (DOA) estimation method based on a combined real-valued subspace in a co-prime matrix, which comprises the steps of firstly improving an array structure of a basic co-prime matrix to ensure that the basic co-prime matrix is centrosymmetric; then, two sub-covariances are constructed through the two sub-arrays, and a relation matrix between the sub-arrays is further obtained; constructing a joint subspace angle of arrival estimation function based on the relation matrix; finally, a DOA estimate is obtained from the intersection of the outputs of the two sub-arrays without ambiguity. Due to the existence of the relation matrix, the output of the two sub-arrays is automatically paired, and the two subspaces are combined, so that the estimation performance is guaranteed. Compared with the current advanced algorithm, the method proposed by the invention greatly reduces the complexity because only real-value feature decomposition and low-dimensional sub-covariance are needed, but can be almost unchanged in DOA estimation performance. Therefore, the invention can greatly reduce the complexity, save the hardware cost and improve the response speed in positioning systems in radar and wireless communication.

Description

Estimation method of angle of arrival based on joint real-value subspace in co-prime matrix
Technical Field
The invention relates to a joint real-value subspace-based angle of arrival estimation method in a co-prime matrix, belonging to array signal processing.
Background
Direction of arrival (DOA) estimation is an important research Direction for array signal processing, and is widely used in many fields such as radar and mobile communication. Many classical super-resolution DOA Estimation methods have been proposed, such as a Multiple signal classification (MUSIC) method, an Estimation of parameters by rotation invariant technology (ESPRIT) method, a propagation operator method (PM) and a rotation invariant PM method. However, to avoid the estimation ambiguity problem, these methods all use compact arrays with array element spacing no greater than half a wavelength, limited DOF results in limited DOA estimation resolution, and interference from mutual coupling between antennas is present.
As a non-uniform sparse array, the nested array can obtain degree of freedom (DOF) far greater than the actual physical array element number in a virtual array domain, so that the system capacity and the spatial resolution are increased, however, a compact sub-array still exists in the nested array, the closer antenna spacing is still influenced by the mutual coupling between the antennas, and the DOA estimation performance is reduced. Another non-uniform sparse array that is currently more popular is a co-prime array, which consists of two uniform linear sub-arrays, where the number of array elements and the spacing of the array elements from each other are mutually prime. The co-prime matrix can also obtain a larger DOF in a virtual domain, and compared with a nested matrix, the array elements of the co-prime matrix are more sparsely distributed, so that the co-prime matrix has stronger robustness to the cross coupling problem. In order to solve the problem of DOA estimation ambiguity caused by large array element spacing, a combined MUSIC method in a document 'DOA estimation with combined MUSIC for coprime array' determines the DOA estimation without ambiguity from the intersection of two sub-array MUSIC spectral peaks based on the co-prime between the sub-arrays of the co-prime array, but the DOA estimation needs two times of global spectral peak search and has high complexity. In order to reduce complexity, a local search MUSIC method in a document 'Partial spectral search-based DOA estimation method for co-prime linear arrays' utilizes the uniformity of a subarray, converts an angle domain estimation problem into a phase domain, can recover all MUSIC peaks of MUSIC by local search, and then solves ambiguity according to mutual homogeneity. However, both joint-MUSIC and local-search-MUSIC process the sub-arrays separately, so that when the final intersection solution is ambiguous, there is interference between multiple DOA estimates (including true solution and ambiguous solution), and especially error is prone at low snr. In order to avoid this problem, the document "Improved DOA estimation algorithm for co-private linear array using root-MUSIC algorithm" applies an Improved root-finding MUSIC method (Li J, junction d. joint evaluation and azimuths estimation for L-shaped array) to the co-prime array, first constructs a relationship matrix of two sub-arrays through the signal subspace of the whole array, and then solves DOA estimation by using a root-finding method based on the noise subspace of the whole array. However, the eigen decomposition of the ensemble of arrays and the estimation function based on the ensemble of noise subspaces both involve a high complexity.
Disclosure of Invention
The invention aims to provide a DOA estimation method based on a joint real-valued subspace in a co-prime matrix. Firstly, the manifold of the array of the co-prime matrix is improved, so that the two sub-matrices are centrosymmetric, and the covariance can be changed into a real value through unitary transformation, thereby greatly reducing the complexity. In addition, the covariance of the whole array is not needed, and the relation matrix between two sub-arrays can be obtained only by the sub-covariance of two real values, so that additional pairing is avoided, and the complexity is further reduced. And finally constructing a DOA estimation cost function of a joint subspace based on the noise subspaces of the two subarrays. Compared with the improved root-finding MUSIC method, the complexity of the algorithm is greatly reduced, but the DOA estimation performance can be almost kept unchanged.
The invention adopts the following technical scheme for solving the technical problems:
the invention provides a method for estimating an angle of arrival based on a joint real-valued subspace in a co-prime matrix, wherein the structures of a first sub-matrix and a second sub-matrix of the co-prime matrix meet central symmetry, and the method comprises the following specific steps:
step 1, constructing a cross covariance between a first subarray and a second subarray and an autocovariance of the second subarray;
step 2, carrying out real-valued transformation on the cross covariance between the first subarray and the second subarray in the step 1 and the auto covariance of the second subarray through unitary transformation;
step 3, carrying out singular value decomposition on the real value transformation result of the cross covariance between the first subarray and the second subarray in the step 2 to obtain a noise subspace corresponding to the first subarray and a noise subspace corresponding to the second subarray;
step 4, constructing a relation matrix according to the real value transformation result of the cross covariance between the first subarray and the second subarray and the auto covariance of the second subarray in the step 2;
step 5, constructing an angle estimation function of the joint real value subspace according to the noise subspace corresponding to the first subarray, the noise subspace corresponding to the second subarray in the step 3 and the relation matrix in the step 4;
step 6, respectively obtaining a group of DOA estimation solutions corresponding to the second subarray and a group of DOA estimation solutions corresponding to the first subarray by utilizing the root solving mode, the solution uniform distribution and the relation matrix;
and 7, taking the average of two solutions closest to each other in the two groups of DOA estimation solutions obtained in the step 6 as a final DOA estimation by utilizing the mutual characteristics of the first subarray and the second subarray.
As a further technical scheme of the invention, in step 1, the cross covariance between the first and second subarrays is
Figure BDA0001687941970000021
The second subarray has an autocovariance of
Figure BDA0001687941970000022
Wherein A is1＝[a1(α1),...,a1(αK)]、A2＝[a2(β1),...,a2(βK)]Respectively, the direction matrixes of the first and second sub-matrixes, K is the number of different directions of the far-field signals existing in the space,
Figure BDA0001687941970000023
αk＝Nπsinθk，βk＝Mπsinθkm, N are the numbers of elements of the first and second sub-arrays, M and N are prime integers, thetakDOA, K-1, 2, …, K,
Figure BDA0001687941970000031
in the form of a diagonal matrix,
Figure BDA0001687941970000032
representing the energy of the k-th signal, σ2As noise energy, INIs an N × N identity matrix.
As a further technical scheme of the invention, in step 2, the real value transformation result of the cross covariance between the first subarray and the second subarray is Rr＝H1RsH2 HThe real value transformation result of the autocovariance of the second subarray is R2r＝H2RsH2 HWherein, in the step (A),
Figure BDA0001687941970000033
are respectively A1、A2Result of unitary transformation of (1), QMAnd QNRepresenting unitary matrices of order M and order N, respectively.
As a further technical solution of the present invention, unitary matrices in even and odd dimensions are defined as:
Figure BDA0001687941970000034
and
Figure BDA0001687941970000035
wherein, IpRepresenting a p x p identity matrix, ΠpDenotes an inverse identity matrix of p × p, p being a natural number.
As a further technical solution of the present invention, step 3 specifically is:
Rris decomposed into Rr＝UΛVTWherein U, Lambda and V are respectively a left singular vector, a singular value and a right singular vector;
noise subspace U corresponding to the first subarraynThe left singular vector corresponding to the minimum M-K singular values is formed;
noise subspace V corresponding to the second subarraynConsisting of right singular vectors corresponding to the smallest N-K singular values.
As a further technical solution of the present invention, in step 4, the relationship matrix G ═ RrR2r +＝H1RsH2 H(H2RsH2 H)+＝H1H2 +Wherein, wherein+Representing the generalized inverse of the matrix.
As a further technical scheme of the invention, the angle estimation function of the joint real-valued subspace in the step 5 is
Figure BDA0001687941970000036
As a further technical solution of the present invention, step 6 specifically is:
solving an angle estimation function of the joint real-valued subspace by using a root solving method, specifically comprising the following steps:
by
Figure BDA0001687941970000037
Is a vandermonde vector, let z equal ejβThe angle estimation function of the joint real-valued subspace is rewritten as
Figure BDA0001687941970000038
a2(z)＝[1,z,…,zN-1]T，
Figure BDA0001687941970000039
Is the union of two noise subspaces;
by means of root finding
Figure BDA00016879419700000310
Solving is carried out, and K roots z which are in the unit circle and are closest to the unit circle are selected from the root finding resultk,k＝1,…,K；
By
Figure BDA0001687941970000041
βk＝MπsinθkObtaining a DOA estimate from the second sub-array
Figure BDA0001687941970000042
Wherein, angle () represents the phase;
obtaining a group of DOA estimation solutions corresponding to the second subarray according to the uniform distribution of all M solutions at intervals of 2/M
Figure BDA0001687941970000043
Wherein the content of the first and second substances,
Figure BDA0001687941970000044
u is an integer which varies with m and is guaranteed
Figure BDA0001687941970000045
In the interval [ -1,1 [)]Within;
by
Figure BDA0001687941970000046
And h1(αk)＝Gh2(βk) To obtain
Figure BDA0001687941970000047
By
Figure BDA0001687941970000048
And is
Figure BDA0001687941970000049
To obtain
Figure BDA00016879419700000410
By a1(αk) The phase difference between adjacent elements can obtain a DOA estimate corresponding to the first sub-array as
Figure BDA00016879419700000411
Wherein a is1,m(αk) And a1,m+1(αk) Respectively represent a1(αk) The m-th and m + 1-th elements of (a);
obtaining a group of DOA estimation solutions corresponding to the first subarray according to the uniform distribution of all N solutions at intervals of 2/N
Figure BDA00016879419700000412
Wherein the content of the first and second substances,
Figure BDA00016879419700000413
v is an integer which varies with n and is guaranteed
Figure BDA00016879419700000414
In the interval [ -1,1 [)]Within.
As a further technical solution of the present invention, the cross covariance between the first and second sub-arrays and the auto covariance of the second sub-array are estimated by a finite number of snapshots:
Figure BDA00016879419700000415
wherein T represents the number of fast beats, x1(t)、x2And (t) the outputs of the first subarray and the second subarray respectively.
Compared with the prior art, the invention adopting the technical scheme has the following technical effects:
1. the improvement of the coprime array structure leads the coprime array structure to be centrosymmetric;
2. based on a symmetrical array structure, the sub-covariance is converted into real values, so that the operation complexity is reduced;
3. the characteristic decomposition and noise space of the whole array are not needed, the relation matrix can be obtained only through the sub-covariance of the real values, additional pairing is avoided, and the complexity is further reduced;
4. finally, constructing a DOA estimation function based on real-value noise subspaces of the two sub-arrays, obtaining DOA estimation information from the two sub-arrays, and performing ambiguity resolution based on the mutual prime;
5. compared with the current advanced algorithm, the proposed algorithm can greatly reduce the complexity, but the DOA estimation performance is almost unchanged.
Drawings
FIG. 1 is a schematic diagram of the basic co-prime array and the improved co-prime array structure;
FIG. 2 is a schematic flow diagram of the method of the present invention;
FIG. 3 is a schematic diagram of the comparison of the algorithm complexity of the improved root MUSIC method and the method of the present invention;
FIG. 4 is the DOA estimation result of the proposed method in the first 100 experiments;
FIG. 5 is a graph illustrating the comparison between the estimation performance of the method of the present invention and the improved root MUSIC algorithm.
Detailed Description
The technical scheme of the invention is further explained in detail by combining the attached drawings:
the invention provides a Direction of arrival (DOA) estimation method based on a joint real-valued subspace in a co-prime matrix. Firstly, the basic co-prime matrix is improved to be centrosymmetric, so that the covariance of the matrix can be changed into a real value through unitary transformation, and the operation complexity is greatly reduced. And then constructing two sub-covariances through the two sub-arrays, and further obtaining a relation matrix between the sub-arrays. Based on the relationship matrix, a joint subspace angle of arrival estimation function is constructed. Finally, a DOA estimate is obtained from the intersection of the outputs of the two sub-arrays without ambiguity. Due to the existence of the relation matrix, the output of the two sub-arrays is automatically paired, and the two subspaces are combined, so that the estimation performance is guaranteed. Compared with the current advanced algorithm, the method proposed by the invention greatly reduces the complexity because only real-value feature decomposition and low-dimensional sub-covariance are needed, but can be almost unchanged in DOA estimation performance. Therefore, the invention can greatly reduce the complexity, save the hardware cost and improve the response speed in positioning systems in radar and wireless communication.
The present co-prime array and the improved co-prime array structure shown in fig. 1, the upper half of fig. 1 shows the array structure of the basic co-prime array, which is composed of two sparse Uniform Linear Arrays (ULA), the sub-array 1 has M array elements with an array element spacing Nd, and the sub-array 2 has N array elements with an array element spacing Md, d being a unit spacing, generally set to be a half wavelength, and M and N being co-prime integers. The array structure does not satisfy central symmetry, so real-valued transformation cannot be performed. We therefore show an improved co-prime array structure in the lower half of fig. 1, so that the two sub-arrays are now symmetric about the Z-axis (it should be noted that fig. 1 shows only an example, and it is only necessary to ensure that the two sub-arrays are symmetric about the Z-axis in practical application of the method of the present invention). Assuming that there are far-field signals from K different directions in the space at this time, the outputs of the two sub-arrays are
x1(t)＝A1s(t)+n1(t) (1)
x2(t)＝A2s(t)+n2(t) (2)
Wherein s (t) ═ s1(t),...,sK(t)]TRepresenting K source baseband signals; n is1(t) and n2(t) represents additive white gaussian noise on the respective arrays; a. the1＝[a1(α1),...,a1(αK)]And A2＝[a2(β1),...,a2(βK)]The direction matrices representing sub-array 1 and sub-array 2, respectively, the columns of which are called steering vectors and are represented as
Figure BDA0001687941970000061
Figure BDA0001687941970000062
Wherein alpha isk＝Nπsinθk,βk＝Mπsinθk,θkIndicating the DOA of the k-th signal.
Real value transformation
First, according to (1) - (2), the cross covariance between two sub-arrays is constructed as
Figure BDA0001687941970000063
Wherein
Figure BDA0001687941970000064
Is a diagonal matrix containing K signal energies (Gu J F, Wei P. Joint SVD of Two Cross-Correlation matrix to arrival Automatic Pairing in 2-D Angle Estimation schemes). Since the two sub-matrices are now centrosymmetric, we can use unitary transform for real-valued transform, thereby reducing complexity. Unitary matrices in odd and even dimensions are defined as follows:
Figure BDA0001687941970000065
Figure BDA0001687941970000066
using the steering vector of sub-matrix 1 as an example, the unitary transformation is then transformed into a real steering vector (assuming M is odd)
Figure BDA0001687941970000067
The corresponding real-valued direction matrix is
Figure BDA0001687941970000068
Similarly, the directional unitary transformation of sub-matrix 2 becomes
Figure BDA0001687941970000069
So that the cross-covariance in equation (5) becomes
Figure BDA00016879419700000610
In practice, to ensure the real value of the cross covariance, we can take the real part of the above equation, i.e. the real part
Figure BDA00016879419700000611
Based on the cross-covariance in equation (8), its Singular Value Decomposition (SVD) is
Rr＝UΛVT (9)
Wherein U, Λ and V represent the left singular vector, the singular value and the right singular vector, respectively. The left singular vector corresponding to the minimum (M-K) singular values constitutes the noise subspace U of subarray 1nAnd the right singular vector corresponding to the minimum (N-K) singular values constitutes the noise subspace V corresponding to sub-array 2n. The angular information of the two sub-arrays can be obtained from the following cost function
Un Hh1(α)＝0 (10)
Vn Hh2(β)＝0 (11)
However, it should be noted that when α and β are obtained separately, then this brings about two problems: 1.α and β are out of order, requiring additional pairing steps, adding complexity; 2. alpha and beta are respectively obtained based on respective sub-arrays, which is equivalent to using only part of the array aperture, so that the estimated performance is to be improved. Therefore, before the angle estimation, we need to obtain the relationship between the two sub-arrays, so as to jointly use the two sub-arrays.
Relationship matrix construction
Constructing subarray 2 with an autocovariance of
Figure BDA0001687941970000071
Wherein sigma2For noise energy, it is also unitary transformed and noise eliminated
Figure BDA0001687941970000072
Wherein sigma2Can be selected from
Figure BDA0001687941970000073
Average of the N-K smaller eigenvalues of ([16 ]]Yunhe C. Joint estimation of angle and Doppler frequency for stationary MIMO radar) it is noted here that the unitary matrix satisfies orthogonality, i.e. it is required to
Figure BDA0001687941970000074
The unitary transform does not have an impact on the noise energy. According to the expressions (8) and (13), a relationship matrix can be constructed as
G＝RrR2r +＝H1RsH2 H(H2RsH2 H)+＝H1H2 + (14)
The G matrix then gives the relation between the two sub-matrix direction matrices, i.e.
H1＝GH2 (15)
It should be noted that the construction of the G matrix is based on the sub-covariances in the equations (8) and (13), and the construction complexity is greatly reduced compared with the covariances of the entire array.
Joint subspace-based angle of arrival estimation
Based on the G matrix, h1(α)＝Gh2(β) into equation (10), then the joint real-valued subspace-based angle estimation function can be constructed as
Figure BDA0001687941970000075
(16) The equation can be solved in a root-finding manner because
Figure BDA0001687941970000081
Wherein
Figure BDA0001687941970000082
Is a vandermonde vector. Let z be ejβThen, then
Figure BDA0001687941970000083
Wherein a is2(z)＝[1,z,…,zN-1]TThen equation (16) can be written as
Figure BDA0001687941970000084
Wherein
Figure BDA0001687941970000085
Is the union of two noise subspaces. Based on the polynomial root finding of equation (17), we select the K roots z in the unit circle and closest to the unit circlek,k＝1,…,K。
Because of the fact that
Figure BDA0001687941970000086
And beta isk＝MπsinθkSo the DOA derived from subarray 2 (i.e., β) is estimated as
Figure BDA0001687941970000087
From the above equation, it can be seen that the sub-array 2 has a larger array element spacing (M)>1)，zkWill exceed [ -pi, pi]So there is a phase ambiguity, we are based on zkThe solution obtained from equation (18) is only one of the fuzzy solutions, and there are (M-1) solutions
Figure BDA0001687941970000088
The M solutions should satisfy
Figure BDA0001687941970000089
So the relationship between them ([11 ]]SUN F, LAN P and GAO B. partial spectral search-based DOA estimation method for co-primary linear arrays) as
Figure BDA00016879419700000810
Where u is an integer which varies with m and is guaranteed
Figure BDA00016879419700000811
In the interval [ -1,1 [)]Within. Therefore, all M solutions have a uniformly distributed relationship and a distance of 2/M, and according to the relationship and one solution in the formula (18), we can recover all M solutions.
We now obtain a set of solutions from subarray 2
Figure BDA00016879419700000812
Next, we need to consider the DOA estimate from sub-array 1. Because of the fact that
Figure BDA00016879419700000813
And
Figure BDA00016879419700000814
and because the relation h is satisfied between the guide vectors1(αk)＝Gh2(βk) Thus obtaining
Figure BDA00016879419700000815
While
Figure BDA00016879419700000816
And is
Figure BDA00016879419700000817
Then
Figure BDA00016879419700000818
Therein
Figure BDA00016879419700000819
Because of the scalar quantity, it can be eliminated by normalization. By a1(αk) The phase difference between adjacent elements can obtain the angle estimation of the corresponding subarray 1
Figure BDA0001687941970000091
Similarly, the above equation is a fuzzy solution, and a set of solutions can be obtained based on the uniform distribution of all N solutions at intervals of 2/N, and one of the solutions obtained by equation (22)
Figure BDA0001687941970000092
We now obtain two sets of solutions for DOA estimation. These two sets of solutions contain many erroneous solutions but also true DOA estimates. According to ZHOU C, SHI Z, GU Y, et al DECOM: DOA estimation with combined MUSIC for coprime array, due to the co-prime in the co-prime array, a non-ambiguous DOA can be obtained from the common solution of the two solutions, i.e. the intersection.
However, in practice, the covariance in equations (5) and (12) can only be estimated by a limited number of snapshots, i.e.
Figure BDA0001687941970000093
Figure BDA0001687941970000094
Where T represents the number of fast beats. Due to the effect of residual noise, there is often no completely coincident solution in the final two sets of solutions, so we can take the average of the two solutions that are closest in the two sets of solutions as the final unambiguous DOA estimate. Because the solutions obtained for (18) and (22) are both based on zkK, so are auto-paired, i.e., both sets of solutions are for the kth source. This is so that the two closest solutions are solved without being affected by the DOAs from the rest.
The invention relates to a method for estimating DOA (Direction of arrival, DOA) based on a joint real-valued subspace in a co-prime matrix, which comprises the following steps of:
1. according to the formulas (23) - (24), two sub-covariances R are constructedcAnd R2；
2. Based on the formula (8) and the formula (13), the two sub-covariances are subjected to real value transformation to obtain RrAnd R2rAnd according to the formula (9) to RrSingular value decomposition is carried out to obtain a noise subspace U corresponding to the subarray 1nAnd noise subspace V of corresponding subarray 2n；
3. By G ═ RrR2r +Obtaining a relation matrix G;
4. according to Un，VnAnd obtaining a joint subspace angle estimation function f (beta) in the formula (16) by using the relation matrix G;
5. obtaining a DOA estimate for the corresponding subarray 2 using a root-finding approach, and obtaining a set of solutions based on equation (19) using the uniform distribution of the solutions
Figure BDA0001687941970000095
6. Obtaining a group of DOA estimation solutions corresponding to the subarray 1 by using the relation matrix in the same way
Figure BDA0001687941970000096
And finally, using the reciprocity as the final DOA estimation from the average of the two solutions which are closest in the two groups of solutions.
The algorithm does not need additional pairing and spectral peak search, and the computational complexity of the algorithm mainly focuses on the construction of the sub-covariance and the solving of the relation matrix and the final root solving. It should be noted that the sub-covariances in the algorithm are all real values, so the computational complexity is only 1/4 for the corresponding complex matrix operation. Eventually, the total number of complex multiplications is required to be about O (MNT + N)2T+0.25*(N3+2MN2) + N) times. The improved root-finding MUSIC method requires the multiplication times of O ((M + N)2T+(M+N)3+MNK+2N2K+K3+N)。
In fig. 3, we show a comparison of two algorithm complexities under a typical parameter configuration, where M is 4, N is 3, K is 2, and T is 100. From fig. 3, it can be seen that the complexity of the algorithm we propose is only 40% of the improved root-finding MUSIC algorithm.
Simulation of experiment
In the experiment, we adopted the proposed improved co-prime array structure, as shown in fig. 1, where M is 4, N is 3, and K is 2 sources, whose DOAs are θ respectively120 ° and θ235 ° is set. Each covariance matrix is estimated by T-100 snapshots. To measure the DOA estimation performance, a Root Mean Square Error (RMSE) is defined as
Figure BDA0001687941970000101
Wherein
Figure BDA0001687941970000102
Is DOA theta at the first Monte Carlo experimentkIn total, we performed 500 experiments.
Fig. 4 shows the DOA estimation result of the proposed algorithm in the previous 100 experiments, and it can be found that the proposed algorithm can always effectively and accurately estimate the DOA of two information sources.
Fig. 5 shows a comparison between the estimated performance of the proposed algorithm and the improved root MUSIC algorithm, where the vertical axis is RMSE and the horizontal axis is SNR, which can be found to be almost coincident, which shows that the proposed algorithm has almost the same performance as the improved root MUSIC algorithm, but the complexity of the proposed algorithm is greatly reduced.
A DOA estimation algorithm based on a joint real-valued subspace in a co-prime matrix is proposed. The key work we do can be summarized as follows:
1. the improvement of the coprime array structure leads the coprime array structure to be centrosymmetric;
2. based on a symmetrical array structure, the sub-covariance is converted into real values, so that the operation complexity is reduced;
3. the characteristic decomposition and noise space of the whole array are not needed, and a relation matrix can be obtained only through the sub-covariance of the real values;
4. finally, constructing a DOA estimation function based on real-value noise subspaces of the two sub-arrays, obtaining DOA estimation information from the two sub-arrays, and performing ambiguity resolution based on the mutual prime;
5. compared with the current advanced algorithm, the proposed algorithm can greatly reduce the complexity, but the DOA estimation performance is almost unchanged.
The above description is only an embodiment of the present invention, but the scope of the present invention is not limited thereto, and any person skilled in the art can understand that the modifications or substitutions within the technical scope of the present invention are included in the scope of the present invention, and therefore, the scope of the present invention should be subject to the protection scope of the claims.

Claims (3)

1. A method for estimating an angle of arrival based on a combined real-valued subspace in a co-prime matrix is characterized in that the structures of a first sub-matrix and a second sub-matrix of the co-prime matrix respectively meet central symmetry, and the method specifically comprises the following steps:
step 1, constructing a cross covariance between a first subarray and a second subarray and an autocovariance of the second subarray; wherein the cross-covariance between the first and second subarrays is
Figure FDA0003363444090000011
The second subarray has an autocovariance of
Figure FDA0003363444090000012
Wherein A is1＝[a1(α1),...,a1(αK)]、A2＝[a2(β1),...,a2(βK)]Respectively, the direction matrixes of the first and second sub-matrixes, K is the number of different directions of the far-field signals existing in the space,
Figure FDA0003363444090000013
Figure FDA0003363444090000014
αk＝Nπsinθk，βk＝Mπsinθkm, N denotes the number of elements in the first and second sub-arrays, thetakDOA, K-1, 2, …, K,
Figure FDA0003363444090000015
in the form of a diagonal matrix,
Figure FDA0003363444090000016
representing the energy of the k-th signal, σ2As noise energy, INAn identity matrix of NxN;
step 2, carrying out real-valued transformation on the cross covariance between the first subarray and the second subarray in the step 1 and the auto covariance of the second subarray through unitary transformation; wherein the real-valued transformation result of the cross-covariance between the first and second sub-arrays is Rr＝H1RsH2 HThe real value transformation result of the autocovariance of the second subarray is R2r＝H2RsH2 HWherein, in the step (A),
Figure FDA0003363444090000017
Figure FDA0003363444090000018
are respectively A1、A2Result of unitary transformation of (1), QMAnd QNUnitary matrices of order M and order N, respectively;
step 3, carrying out singular value decomposition on the real value transformation result of the cross covariance between the first subarray and the second subarray in the step 2 to obtain a noise subspace corresponding to the first subarray and a noise subspace corresponding to the second subarray; the method specifically comprises the following steps:
Rris decomposed into Rr＝UΛVTWherein U, Lambda and V are respectively a left singular vector, a singular value and a right singular vector;
noise subspace U corresponding to the first subarraynThe left singular vector corresponding to the minimum M-K singular values is formed;
noise subspace V corresponding to the second subarraynRight singular vectors corresponding to the minimum N-K singular values;
step 4, constructing a relation matrix according to the real value transformation result of the cross covariance between the first subarray and the second subarray and the auto covariance of the second subarray in the step 2: g ═ RrR2r +＝H1RsH2 H(H2RsH2 H)+＝H1H2 +Wherein, wherein+Representing a generalized inverse of the matrix;
step 5, constructing an angle estimation function of the joint real value subspace according to the noise subspace corresponding to the first subarray, the noise subspace corresponding to the second subarray in the step 3 and the relation matrix in the step 4
Figure FDA0003363444090000019
Step 6, respectively obtaining a group of DOA estimation solutions corresponding to the second subarray and a group of DOA estimation solutions corresponding to the first subarray by utilizing the root solving mode, the solution uniform distribution and the relation matrix; the method specifically comprises the following steps:
solving an angle estimation function of the joint real-valued subspace by using a root solving method, specifically comprising the following steps:
by
Figure FDA0003363444090000021
Is a vandermonde vector, let z equal ejβThe angle estimation function of the joint real-valued subspace is rewritten as
Figure FDA0003363444090000022
a2(z)＝[1,z,…,zN-1]T，
Figure FDA0003363444090000023
Is the union of two noise subspaces;
by means of root finding
Figure FDA0003363444090000024
Solving is carried out, and K roots z which are in the unit circle and are closest to the unit circle are selected from the root finding resultk,k＝1,…,K；
By
Figure FDA0003363444090000025
βk＝MπsinθkObtaining a DOA estimate from the second sub-array
Figure FDA0003363444090000026
Wherein, angle () represents the phase;
obtaining a group of DOA estimation solutions corresponding to the second subarray according to the uniform distribution of all M solutions at intervals of 2/M
Figure FDA0003363444090000027
Wherein the content of the first and second substances,
Figure FDA0003363444090000028
u is an integer which varies with m and is guaranteed
Figure FDA0003363444090000029
In the interval [ -1,1 [)]Within;
by
Figure FDA00033634440900000210
And h1(αk)＝Gh2(βk) To obtain
Figure FDA00033634440900000211
By
Figure FDA00033634440900000212
And is
Figure FDA00033634440900000213
To obtain
Figure FDA00033634440900000214
By a1(αk) The phase difference between adjacent elements can obtain a DOA estimate corresponding to the first sub-array as
Figure FDA00033634440900000215
Wherein a is1,m(αk) And a1,m+1(αk) Respectively represent a1(αk) The m-th and m + 1-th elements of (a);
obtaining a group of DOA estimation solutions corresponding to the first subarray according to the uniform distribution of all N solutions at intervals of 2/N
Figure FDA00033634440900000216
Wherein the content of the first and second substances,
Figure FDA00033634440900000217
v is an integer which varies with n and is guaranteed
Figure FDA00033634440900000218
In the interval [ -1,1 [)]Within;
and 7, taking the average of two solutions closest to each other in the two groups of DOA estimation solutions obtained in the step 6 as a final DOA estimation by utilizing the mutual characteristics of the first subarray and the second subarray.
2. The method of claim 1, wherein the unitary matrices in even and odd dimensions are defined as:
Figure FDA00033634440900000219
and
Figure FDA00033634440900000220
wherein, IpRepresenting a p x p identity matrix, ΠpDenotes an inverse identity matrix of p × p, p being a natural number.
3. The joint real-valued subspace-based angle of arrival estimation method in a co-prime matrix of claim 1, wherein the cross-covariance between the first and second sub-matrices and the auto-covariance of the second sub-matrix are estimated by finite snapshot numbers:
Figure FDA0003363444090000031
wherein T represents the number of fast beats, x1(t)、x2And (t) the outputs of the first subarray and the second subarray respectively.
CN201810579211.3A 2018-06-07 2018-06-07 Estimation method of angle of arrival based on joint real-value subspace in co-prime matrix Active CN108896954B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201810579211.3A CN108896954B (en) 2018-06-07 2018-06-07 Estimation method of angle of arrival based on joint real-value subspace in co-prime matrix

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201810579211.3A CN108896954B (en) 2018-06-07 2018-06-07 Estimation method of angle of arrival based on joint real-value subspace in co-prime matrix

Publications (2)

Publication Number Publication Date
CN108896954A CN108896954A (en) 2018-11-27
CN108896954B true CN108896954B (en) 2022-01-11

Family

ID=64344093

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201810579211.3A Active CN108896954B (en) 2018-06-07 2018-06-07 Estimation method of angle of arrival based on joint real-value subspace in co-prime matrix

Country Status (1)

Country Link
CN (1) CN108896954B (en)

Families Citing this family (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109613504A (en) * 2018-12-07 2019-04-12 中国人民解放军空军工程大学 A kind of quick angle estimation method of sparse linear array
CN110187304B (en) * 2019-05-21 2021-05-04 泰凌微电子(上海)股份有限公司 Signal arrival angle estimation method and device
CN110297213A (en) * 2019-07-17 2019-10-01 南京航空航天大学 Radiation source positioning device and method based on the unmanned aerial vehicle platform for loading relatively prime linear array
CN110749858B (en) * 2019-09-26 2023-07-21 南京航空航天大学 Polynomial root finding based direction finding estimation method for unfolding mutual mass array
CN110927658A (en) * 2019-12-04 2020-03-27 南京理工大学实验小学 Method for optimizing reciprocity number in reciprocity linear array
CN111025225A (en) * 2019-12-18 2020-04-17 南京航空航天大学 Propagation operator-based direction of arrival estimation method suitable for co-prime linear array
CN111352063B (en) * 2019-12-20 2023-08-01 南京航空航天大学 Two-dimensional direction finding estimation method based on polynomial root finding in uniform area array
CN113219398B (en) * 2020-06-22 2022-09-13 哈尔滨工业大学（威海） Far-field narrow-band radio signal direction-of-arrival estimation method
CN111965590A (en) * 2020-06-30 2020-11-20 南京航空航天大学 Low-complexity DOA estimation method based on co-prime circular array
CN113219399B (en) * 2020-08-05 2023-03-10 哈尔滨工业大学（威海） Far-field narrow-band radio signal direction-of-arrival estimation method based on full real value calculation

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102213761A (en) * 2011-04-06 2011-10-12 哈尔滨工程大学 Multi-target location method of bistatic common-address multi-input-multi-output radar
CN106019213A (en) * 2016-05-09 2016-10-12 电子科技大学 Partial sparse L array and two-dimensional DOA estimation method thereof
CN106291451A (en) * 2016-08-17 2017-01-04 河海大学 DoA method of estimation based on multiple signal classification group delay algorithm
CN106802402A (en) * 2017-03-09 2017-06-06 西安电子科技大学 DOA estimation method based on dual-layer Parallel circular array antenna
CN106980106A (en) * 2017-04-21 2017-07-25 天津大学 Sparse DOA estimation method under array element mutual coupling
CN107589399A (en) * 2017-08-24 2018-01-16 浙江大学 Based on the relatively prime array Wave arrival direction estimating method for sampling virtual signal singular values decomposition more
CN107894581A (en) * 2017-11-16 2018-04-10 河海大学 A kind of wideband array Wave arrival direction estimating method

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102213761A (en) * 2011-04-06 2011-10-12 哈尔滨工程大学 Multi-target location method of bistatic common-address multi-input-multi-output radar
CN106019213A (en) * 2016-05-09 2016-10-12 电子科技大学 Partial sparse L array and two-dimensional DOA estimation method thereof
CN106291451A (en) * 2016-08-17 2017-01-04 河海大学 DoA method of estimation based on multiple signal classification group delay algorithm
CN106802402A (en) * 2017-03-09 2017-06-06 西安电子科技大学 DOA estimation method based on dual-layer Parallel circular array antenna
CN106980106A (en) * 2017-04-21 2017-07-25 天津大学 Sparse DOA estimation method under array element mutual coupling
CN107589399A (en) * 2017-08-24 2018-01-16 浙江大学 Based on the relatively prime array Wave arrival direction estimating method for sampling virtual signal singular values decomposition more
CN107894581A (en) * 2017-11-16 2018-04-10 河海大学 A kind of wideband array Wave arrival direction estimating method

Also Published As

Publication number Publication date
CN108896954A (en) 2018-11-27

Similar Documents

Publication Publication Date Title
CN108896954B (en) Estimation method of angle of arrival based on joint real-value subspace in co-prime matrix
CN107092004B (en) Estimation method of direction of arrival of co-prime array based on signal subspace rotation invariance
CN109490820B (en) Two-dimensional DOA estimation method based on parallel nested array
CN110045323B (en) Matrix filling-based co-prime matrix robust adaptive beamforming algorithm
Yan et al. Fast DOA estimation based on a split subspace decomposition on the array covariance matrix
CN108120967B (en) Plane array DOA estimation method and equipment
CN110161452B (en) Direction-of-arrival estimation method based on cross-prime L-shaped electromagnetic vector sensor array
CN110244272B (en) Direction-of-arrival estimation method based on rank-denoising model
CN112379327A (en) Two-dimensional DOA estimation and cross coupling correction method based on rank loss estimation
CN111965598B (en) Method for positioning radiation source by utilizing non-uniform array
CN110531312B (en) DOA estimation method and system based on sparse symmetric array
Gu et al. Design of wide scanning sparse planar array using both matrix-pencil and space-mapping methods
CN114239644A (en) Non-grid block sparse angle estimation method based on array element cross coupling optimization
CN113189592A (en) Vehicle-mounted millimeter wave MIMO radar angle measurement method considering amplitude mutual coupling error
Li et al. Sparse nested linear array for direction of arrival estimation
CN113075610B (en) DOA estimation method for differential array interpolation based on co-prime polarization array
Tian et al. 2D-DOA estimation in arc-array with a DNN based covariance matrix completion strategy
He et al. Efficient underwater two-dimensional coherent source localization with linear vector-hydrophone array
Liu et al. Two-dimensional doa estimation via matrix completion and sparse matrix recovery for coprime planar array
CN114487990A (en) Two-dimensional DOA estimation method, device, equipment and medium based on parallel nested array
Li et al. Fast direction of arrival estimation using a sensor-saving coprime array with enlarged inter-element spacing
CN113791379A (en) Orthogonal matching pursuit DOA estimation method under nested array non-Gaussian environment
Hui et al. An improved algorithm of ESPRIT for signal DOA estimation
Wang et al. A two-dimensional direction finding method based on non-uniform array
CN112698263A (en) Orthogonal propagation operator-based single-basis co-prime MIMO array DOA estimation algorithm

Legal Events

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