High-resolution, accurate, two-dimensional direction-of-arrival estimation method based on coarray tensor spatial spectrum searching with co-prime planar array

Abstract

Disclosed is a high-resolution accurate two-dimensional direction-of-arrival estimation method based on coarray tensor spatial spectrum searching with coprime planar array, which solves the problem of multi-dimensional signal loss and limited spatial spectrum resolution and accuracy in existing methods. The implementation steps are: constructing a coprime planar array; tensor signal modeling for the coprime planar array; deriving coarray statistics based on coprime planar array cross-correlation tensor; constructing the equivalent signals of a virtual uniform array; deriving a spatially smoothed fourth-order auto-correlation coarray tensor; realizing signal and noise subspace classification through coarray tensor feature extraction; performing high-resolution accurate two-dimensional direction-of-arrival estimation based on coarray tensor spatial spectrum searching. In the present method, multi-dimensional feature extraction based on coarray tensor statistics for coprime planar array is used to implement high-resolution, accurate two-dimensional direction-of-arrival estimation based on tensor spatial spectrum searching, and the method can be used for passive detection and target positioning.

Claims

1. A high-resolution accurate two-dimensional direction-of-arrival estimation method based on coarray tensor spatial spectrum searching with coprime planar array, comprising the following steps of: (1) providing a receiving end, which is constituted by 4M.sub.xM.sub.y+N.sub.xN.sub.y−1 physical antenna sensors arranged in a coprime planar array, wherein M.sub.x, N.sub.x and M.sub.y, N.sub.y are respectively a pair of prime integers, and M.sub.x<N.sub.x, M.sub.y<N.sub.y; and the receiving end is decomposed into two sparse uniform subarrays, which are respectively a first sparse uniform subarray .sub.1 and a second sparse uniform subarray .sub.2; (2) receiving, by the receiving end, signals of K far-field narrowband incoherent sources from directions of {(θ.sub.1, φ.sub.1), (θ.sub.2, φ.sub.2), . . . , (θ.sub.K, φ.sub.K)}; and processing the signals as the following manner; the received signals of the first sparse subarray .sub.1 being expressed by using a three-dimensional tensor .sub.1∈.sup.2M.sup.x.sup.×2M.sup.y.sup.×L (L is the number of snapshots) as: $x_1=\underset{k=1}{\overset{K}{.Math.}}{a}_{Mx}\left({\theta }_k,{\phi }_k\right)\circ a_{\mathrm{My}}\left({\theta }_k,{\phi }_k\right)\circ s_k+{𝒩}_1,$ where s.sub.k=[.sub.k,1, s.sub.k,2, . . . , s.sub.k,L].sup.T is a multi-snapshot sampling signal waveform corresponding to the k.sup.th incident source, [⋅].sup.T represents transposition operation, ∘ represents a vector outer product, .sub.1 is a noise tensor independent of each source, a.sub.Mx(θ.sub.k, φ.sub.k) and a.sub.My(θ.sub.k, φ.sub.k) are respectively steering vectors of the first sparse uniform subarray .sub.1 in x-axis and y-axis directions, corresponding to the k.sup.th source with a direction-of-arrival of (θ.sub.k, φ.sub.k), which are expressed as: $a_{\mathrm{Mx}}\left({\theta }_k,{\phi }_k\right)={\left[1,e^{-j\pi u_1^{\left(2\right)}\sin \left({\phi }_k\right)\cos \left({\theta }_k\right)},\mathrm{.Math.},e^{-j\pi u_1^{\left(2M_x\right)}\sin \left({\phi }_k\right)\cos \left({\theta }_k\right)}\right]}^T,a_{\mathrm{My}}\left({\theta }_k,{\phi }_k\right)={\left[1,e^{-j\pi v_1^{\left(2\right)}\sin \left({\phi }_k\right)\cos \left({\theta }_k\right)},\mathrm{.Math.},e^{-j\pi v_1^{\left(2M_y\right)}\sin \left({\phi }_k\right)\cos \left({\theta }_k\right)}\right]}^T,$ where u.sub.1.sup.(i.sup.1.sup.) (i.sub.1=1, 2, . . . , 2M.sub.x) and .sup.(i.sup.2.sup.) (i.sub.2=1, 2, . . . , 2M.sub.y) respectively represent actual positions of the i.sub.1.sup.th physical sensor and the i.sub.2.sup.th physical sensor of the first sparse uniform subarray .sub.1 in the x-axis and y-axis directions, and u.sub.1.sup.(1)=0, .sub.1.sup.(1)=0, j=√{square root over (−1)}; and the received signals of the sparse subarray .sub.2 being expressed by using another three-dimensional tensor .sub.2∈.sup.N.sup.x.sup.×N.sup.y.sup.×L as: $x_2=\underset{k=1}{\overset{K}{.Math.}}{a}_{\mathrm{Nx}}\left({\theta }_k,{\phi }_k\right)\circ a_{\mathrm{Ny}}\left({\theta }_k,{\phi }_k\right)\circ s_k+{𝒩}_2,$ where .sub.2 is a noise tensor independent of each source, a.sub.Nx(θ.sub.k, φ.sub.k) and a.sub.Ny(θ.sub.k, φ.sub.k) are respectively steering vectors of the second sparse uniform subarray .sub.2 in the X-axis and Y-axis directions, corresponding to the k.sup.th source with a direction-of-arrival of (θ.sub.k, φ.sub.k) which are expressed as: $a_{\mathrm{Nx}}\left({\theta }_k,{\phi }_k\right)={\left[1,e^{-j\pi u_2^{\left(2\right)}\sin \left({\phi }_k\right)\cos \left({\theta }_k\right)},\mathrm{.Math.},e^{-j\pi u_2^{\left(N_x\right)}\sin \left({\phi }_k\right)\cos \left({\theta }_k\right)}\right]}^T,a_{\mathrm{Ny}}\left({\theta }_k,{\phi }_k\right)={\left[1,e^{-j\pi v_2^{\left(2\right)}\sin \left({\phi }_k\right)\cos \left({\theta }_k\right)},\mathrm{.Math.},e^{-j\pi v_2^{\left(N_y\right)}\sin \left({\phi }_k\right)\cos \left({\theta }_k\right)}\right]}^T,$ where u.sub.2.sup.(i.sup.3.sup.) (i.sub.3=1, 2, . . . , N.sub.x) and .sub.2.sup.(i.sup.4.sup.) (i.sub.4=1, 2, . . . , N.sub.y) respectively represent actual positions of the i.sub.3.sup.th physical sensor and the i.sub.4.sup.th physical sensor of the second sparse uniform subarray .sub.2 in the x-axis and y-axis directions, and u.sub.2.sup.(1)=0, .sub.2.sup.(1)=0; a second-order cross-correlation tensor ∈.sup.2M.sup.x.sup.×2M.sup.y.sup.×N.sup.x.sup.×N.sup.y of the received tensor signals .sub.1 and .sub.2 of the first-sparse uniform subarray .sub.1 and the second sparse uniform subarray .sub.2 is calculated as follows: $\hat{R}=\frac{1}{L}\underset{l=1}{\stackrel{L}{.Math.}}{x}_1\left(l\right)\circ x_2^{*}\left(l\right),$ where, .sub.1(l) and .sub.2(l) respectively represent a l.sup.th slice of .sub.1 and .sub.2 in a third dimension (i.e., a snapshot dimension), and (⋅)* represents a conjugate operation; (3) obtaining, by the receiving end, an augmented non-uniform virtual array based on the cross-correlation tensor , wherein every element in PARAFAC-based unfolding of the cross-correlation tensor corresponds to a virtual sensor, and a position of each virtual sensor is expressed as:
={(−M.sub.xn.sub.xd+N.sub.xm.sub.xd,−M.sub.yn.sub.yd+N.sub.ym.sub.yd)|0≤n.sub.x≤N.sub.x−1,0≤m.sub.x≤2M.sub.x−1,0≤n.sub.y≤N.sub.y−1,0≤m.sub.y≤2M.sub.y−1}, where a unit interval d is half of a wavelength λ of an incident narrowband signal, that is, d=λ/2; dimension sets .sub.1={1, 3} and ={2, 4} are defined, and then an ideal value (a noise-free scene) of the cross-correlation tensor is subjected to modulo {.sub.1, .sub.2} PARAFAC-based unfolding to obtain an ideal expression of the equivalent signals V∈.sup.2M.sup.x.sup.N.sup.x.sup.×2M.sup.y.sup.N.sup.y of the augmented virtual array as:
V=Σ.sub.k=1.sup.Kσ.sub.k.sup.1a.sub.x(θ.sub.k,φ.sub.k)∘a.sub.y(θ.sub.k,φ.sub.k), where a.sub.x(θ.sub.k, φ.sub.k)=a*.sub.Nx(θ.sub.k, φ.sub.k).Math.a.sub.Mx(θ.sub.k, φ.sub.k) and a.sub.y(θ.sub.k, φ.sub.k)=a*.sub.Ny(θ.sub.k, φ.sub.k).Math.a.sub.My(θ.sub.k, φ.sub.k) are steering vectors of the augmented virtual array in the x-axis and y-axis directions, corresponding to k.sup.th source with a direction-of-arrival of (θ.sub.k, φ.sub.k); σ.sub.k.sup.2 represents the power of a k.sup.th incident source; .Math. represents a Kronecker product; a subscript of the tensor represents the PARAFAC-based tensor unfolding; (4) obtaining, by the receiving end, equivalent signals V of the augmented virtual array , wherein a process of said obtaining the by the receiving end, equivalent signals V of the augmented virtual array comprises: containing a virtual array with an x-axis distribution from (−N.sub.x+1)d to (M.sub.xN.sub.x+M.sub.x−1)d and a y-axis distribution from (−N.sub.y+1)d to (M.sub.yN.sub.y+M.sub.y−1)d; wherein there are D.sub.x×D.sub.y virtual sensors in total in , where D.sub.x=M.sub.xN.sub.x+M.sub.x+N.sub.x−1, D.sub.y=m.sub.yN.sub.y+M.sub.y+N.sub.y−1, and is expressed as:
={(x,y)|x=p.sub.xd,y=p.sub.yd,−N.sub.x+1≤p.sub.x≤M.sub.xN.sub.x+M.sub.x−1,−N.sub.y+1≤p.sub.y≤M.sub.yN.sub.y+M.sub.y−1}, obtaining, by the receiving end, the equivalent signals V∈.sup.D.sup.x.sup.×D.sup.y of the virtual array by selecting elements in the coarray signals V corresponding to the position of each virtual sensor of , wherein the equivalent signals V are expressed as: $\overline{V}=\underset{k=1}{\overset{K}{.Math.}}{\sigma }_k^2{b}_x\left({\theta }_k,{\phi }_k\right)\circ b_y\left({\theta }_k,{\phi }_k\right),$
where
b.sub.x(θ.sub.k,φ.sub.k)=[e.sup.−jπ(−N.sup.x.sup.+1)sin(φ.sup.k.sup.)cos(θ.sup.k.sup.),e.sup.−jπ(−N.sup.x.sup.+2)sin(φ.sup.k.sup.)cos(θ.sup.k.sup.), . . . ,e.sup.−jπ(M.sup.x.sup.N.sup.x.sup.+M.sup.x.sup.−1)sin(φ.sup.k.sup.)cos(θ.sup.k.sup.)]
and
b.sub.y(θ.sub.k,φ.sub.k)=[e.sup.−jπ(−N.sup.y.sup.+1)sin(φ.sup.k.sup.)sin(θ.sup.k.sup.),e.sup.−jπ(−N.sup.y.sup.+2)sin(φ.sup.k.sup.)sin(θ.sup.k.sup.), . . . ,e.sup.−jπ(M.sup.y.sup.N.sup.y.sup.+M.sup.y.sup.−1)sin(φ.sup.k.sup.)sin(θ.sup.k.sup.)] are steering vectors of the virtual array in the x-axis and y-axis directions, corresponding to k.sup.th source with a direction-of-arrival of (θ.sub.k, φ.sub.k); (5) obtaining, by the receiving end, spatially smoothed coarray signals {tilde over (V)} by diving the virtual array , wherein a process of said obtaining, by the receiving end, spatially smoothed coarray signals {tilde over (V)} by diving the virtual array comprises: in the virtual array , taking a subarray with a size of Y.sub.1×Y.sub.2 for every other sensors along the x-axis and y-axis directions respectively, to divide the virtual array into L.sub.1×L.sub.2 uniform subarrays partly overlapping with each other; expressing the above subarray as .sub.(g.sub.1.sub.,g.sub.2.sub.), g.sub.1=1, 2, . . . , L.sub.1, g.sub.2=1, 2, . . . , L.sub.2, and obtaining, by the receiving end, the equivalent signals V.sub.(g.sub.1.sub.,g.sub.2.sub.)∈.sup.Y.sup.1.sup.×Y.sup.2 of the virtual subarray .sub.(g.sub.1.sub.,g.sub.2.sub.) according to respective position elements in the coarray signals V corresponding to the subarray .sub.(g.sub.1.sub.,g.sub.2.sub.):
V.sub.(g.sub.1.sub.,g.sub.2.sub.)=Σ.sub.k=1.sup.Kσ.sub.k.sup.2(c.sub.x(θ.sub.k,φ.sub.k)e.sup.(g.sup.1.sup.−1)sin(φ.sup.k.sup.)cos(θ.sup.k.sup.))∘(c.sub.y(θ.sub.kφ.sub.k)e.sup.(g.sup.2.sup.−1)sin(φ.sup.k.sup.)sin(θ.sup.k.sup.)),
where
c.sub.x(θ.sub.k,φ.sub.k)=[e.sup.−jπ(−N.sup.x.sup.+1)sin(φ.sup.k.sup.)cos(θ.sup.k.sup.),e.sup.−jπ(−N.sup.x.sup.+2)sin(φ.sup.k.sup.)cos(θ.sup.k.sup.), . . . ,e.sup.−jπ(M.sup.x.sup.N.sup.x.sup.+M.sup.x.sup.−1)sin(φ.sup.k.sup.)cos(θ.sup.k.sup.)]
c.sub.y(θ.sub.k,φ.sub.k)=[e.sup.−jπ(−N.sup.y.sup.+1)sin(φ.sup.k.sup.)sin(θ.sup.k.sup.),e.sup.−jπ(−N.sup.y.sup.+2)sin(φ.sup.k.sup.)sin(θ.sup.k.sup.), . . . ,e.sup.−jπ(M.sup.y.sup.N.sup.y.sup.+M.sup.y.sup.−1)sin(φ.sup.k.sup.)sin(θ.sup.k.sup.)] are steering vectors of the virtual subarray .sub.(1,1) in the x-axis and y-axis corresponding to the direction of (θ.sub.k, φ.sub.k); after the above operation, a total of L.sub.1×L.sub.2 coarray signals V.sub.(g.sub.1.sub.,g.sub.2.sub.), the dimensions of which are all Y.sub.1×Y.sub.2 are obtained by the receiving end; and the spatially smoothed coarray signals {tilde over (V)}∈.sup.Y.sup.1.sup.×Y.sup.2.sup.×Y.sup.1.sup.×Y.sup.2 is obtained by the receiving end by taking an average of the L.sub.1×L.sub.2 coarray signals VV.sub.(g.sub.1.sub.,g.sub.2.sub.): $\tilde{V}=\frac{1}{L_1{L}_2}\underset{p=1}{\overset{L_1}{.Math.}}\underset{q=1}{\overset{L_2}{.Math.}}{\overline{V}}_{\left(p,q\right)},$ a fourth-order auto-correlation tensor ∈.sup.Y.sup.1.sup.×Y.sup.2.sup.×Y.sup.1.sup.×Y.sup.2 for the coarray signal {tilde over (V)} calculated by the receiving end as follows:
={tilde over (V)}∘{tilde over (V)}*; (6) performing, by the receiving end, CANDECOMP/PARACFAC decomposition on the fourth-order auto-correlation tensor to extract multi-dimensional features, a result of which is expressed as follows:
=Σ.sub.k=1.sup.K{tilde over (c)}.sub.x(θ.sub.k,φk)∘{tilde over (c)}.sub.y(θ.sub.k,φ.sub.k)∘{tilde over (c)}*.sub.x(θ.sub.k,φ.sub.k)∘{tilde over (c)}*.sub.y(θ.sub.k,φ.sub.k), where {tilde over (c)}.sub.x(θ.sub.k, φ.sub.k)(k=1, 2, . . . , K) and {tilde over (c)}.sub.y(θ.sub.k, φ.sub.k)(k=1, 2, . . . , K) are two orthogonal factor vectors obtained by the CANDECOMP/PARACFAC decomposition, respectively representing spatial information in the x-axis and y-axis directions; C.sub.x=[{tilde over (c)}.sub.x(θ.sub.1, φ.sub.1), {tilde over (c)}.sub.x(θ.sub.2, φ.sub.2), . . . , {tilde over (c)}.sub.x(θ.sub.K, φ.sub.K)] and C.sub.y=[{tilde over (c)}.sub.y(θ.sub.1, φ.sub.1), {tilde over (c)}.sub.y(θ.sub.2, φ.sub.2), . . . , {tilde over (c)}.sub.y(θ.sub.K, φ.sub.K)] are factor matrixes; a space expanded by {{tilde over (c)}.sub.x(θ.sub.k, φ.sub.k)∘{tilde over (c)}.sub.y(θ.sub.k, φ.sub.k), k=1, 2 . . . , K} is taken and is recorded as span{{tilde over (c)}.sub.x(θ.sub.k, φ.sub.k)∘{tilde over (c)}.sub.y(θ.sub.k, φ.sub.k), k=1, 2, . . . , K} as a signal subspace; the signal subspace is expressed using a tensor .sub.s∈.sup.Y.sup.1.sup.×Y.sup.2.sup.×K where .sub.s(k) represents a k.sup.th slice of .sub.s along the third dimension, which is expressed as:
.sub.s(k)={tilde over (c)}.sub.x(θ.sub.k,φ.sub.k)∘{tilde over (c)}.sub.y(θ.sub.k,φ.sub.k); in order to obtain the noise subspace, the orthocomplements of the factor matrices C.sub.x and C.sub.y are calculated; the orthocomplement of C.sub.x is recorded as span{{tilde over (d)}.sub.x,h, h=1, 2, . . . , min(Y.sub.1, Y.sub.2)−K}, and the orthocomplement of C.sub.y is recorded as span{{tilde over (d)}.sub.y,h, h=1, 2, . . . , min(Y.sub.1, Y.sub.2)−K}, where min(⋅) represents the minimum operation; then span{{tilde over (d)}.sub.x,h∘{tilde over (d)}.sub.y,h, h=1, 2, . . . , min(Y.sub.1, Y.sub.2)−K} is taken as the noise subspace, and the tensor .sub.n∈.sup.Y.sup.1.sup.×Y.sup.2.sup.×(min(Y.sup.1.sup.,Y.sup.2.sup.)−K) is used to express the noise subspace, and .sub.n(h) represents a h.sup.th slice of .sub.n along the third dimension, which is expressed as:
.sub.n(h)={tilde over (d)}.sub.x,h∘{tilde over (d)}.sub.y,h; and (7) obtaining, by the receiving end, a two-dimensional direction-of-arrival by tensor spatial spectrum searching, wherein a process of said obtaining, by the receiving end, a two-dimensional direction-of-arrival by tensor spatial spectrum searching comprises: defining a two-dimensional direction-of-arrival ({tilde over (θ)}, {tilde over (φ)}) for spectral peak searching, a {tilde over (θ)}∈[−90°, 90°], {tilde over (φ)}∈[0°, 180° ], and constructing steering information F({tilde over (θ)}, {tilde over (φ)})∈.sup.Y.sup.1.sup.×Y.sup.2 corresponding to the virtual array are constructed, which is expressed as:
F({tilde over (θ)},{tilde over (φ)})=c.sub.x({tilde over (θ)},{tilde over (φ)})∘c.sub.y({tilde over (θ)},{tilde over (φ)}), constructing, by the receiving end, a tensor spatial spectrum function .sub.CP({tilde over (θ)}, {tilde over (φ)}) using the noise subspace obtained from the CANDECOMP/PARACFAC decomposition, which is expressed as follows:
.sub.CP({tilde over (θ)},{tilde over (φ)})=∥<.sub.n×.sub.{1,2}F({tilde over (θ)},{tilde over (φ)})>∥.sub.F.sup.−2, where <x.sub.{Q}> represents a modulo {Q} contraction operation of two tensors along a Q.sup.th dimension, which requires a same size of the Q.sup.th dimension of the two tensors; ∥⋅∥.sub.F represents a Frobenius norm; .sub.n∈.sup.Y.sup.1.sup.×Y.sup.2.sup.×(min(Y.sup.1.sup.,Y.sup.2.sup.)−K) and F({tilde over (θ)}, {tilde over (φ)})∈.sup.Y.sup.1.sup.×Y.sup.2 are subjected to modulo {1, 2} contraction operation along the 1.sup.st and 2.sup.nd dimensions to obtain one vector p∈.sup.min(Y.sup.1.sup.,Y.sup.2.sup.)−K; after obtaining the spatial spectrum function .sub.CP({tilde over (θ)}, {tilde over (φ)}), the receiving end construct the spatial spectrum corresponding to searching directions of the two-dimensional direction-of-arrival, then the estimation of two-dimensional direction-of-arrival of the incident source is obtained by searching for the two-dimensional direction-of-arrival corresponding to the position of the spectral peak.

2. The high-resolution accurate two-dimensional direction-of-arrival estimation method based on coarray tensor spatial spectrum searching with coprime planar array according to claim 1, wherein the receiving end described in step (1) is decomposed into a first sparse uniform subarray .sub.1 and a second sparse uniform subarray .sub.2 constructed on a planar coordinate system xoy, wherein .sub.1 contains 2M.sub.x×2M.sub.y antenna sensors, the sensor spacing in the x-axis direction and the y-axis direction are N.sub.xd and N.sub.yd respectively, and the coordinate of which on xoy is {(N.sub.xdm.sub.x, N.sub.ydm.sub.y), m.sub.x=0, 1, . . . , 2M.sub.x−1, m.sub.y=0, 1, . . . , 2M.sub.y−1}; .sub.2 contains N.sub.x×N.sub.y antenna sensors, the sensor spacing in the x-axis direction and the y-axis direction are M.sub.xd and M.sub.yd respectively, and the coordinate of which on xoy is {(M.sub.xdn.sub.x,M.sub.ydn.sub.y), n.sub.x=0, 1, . . . , N.sub.x−1,n.sub.y=0, 1, . . . , N.sub.y−1}; wherein M.sub.x,N.sub.x and M.sub.y,N.sub.y are respectively a pair of coprime integers, and M.sub.x<N.sub.x,M.sub.y<N.sub.y; .sub.1 and .sub.2 are subjected to subarray combination in a way of overlapping array elements at (0,0) coordinate to obtain a coprime planar array actually containing (4M.sub.xM.sub.y+N.sub.xN.sub.y−1) physical antenna sensors.

3. The high-resolution accurate two-dimensional direction-of-arrival estimation method based on coarray tensor spatial spectrum searching with coprime planar array according to claim 1, wherein the cross-correlation tensor described in step (3) is ideally (a noise-free scene) modeled as:
=Σ.sub.k=1.sup.Kσ.sub.k.sup.2a.sub.Mx(θ.sub.k,φ.sub.k)∘a.sub.My(θ.sub.k,φ.sub.k)∘a*.sub.Nx(θ.sub.k,φ.sub.k)∘a*.sub.Ny(θ.sub.k,φ.sub.k), where a.sub.Mx(θ.sub.k, φ.sub.k)∘a*.sub.Nx(θ.sub.k, φ.sub.k) in is equivalent to an augmented coarray along the x-axis, and a.sub.My(θ.sub.k, φ.sub.k)∘a*.sub.Ny(θ.sub.k, φ.sub.k) is equivalent to an augmented coarray along the y-axis, so that the non-uniform virtual array is obtained.

4. The high-resolution accurate two-dimensional direction-of-arrival estimation method based on coarray tensor spatial spectrum searching with coprime planar array according to claim 1, wherein the equivalent signals V of the virtual uniform array described in step (5) saves spatial structural information of the virtual domain, however, since V can be regarded as the single snapshot coarray signals, the statistics thereof has a rank deficiency problem; therefore, based on the idea of two-dimensional spatial smoothing, the coarray signals V are processed to construct equivalent multi-snapshot coarray signals; after the partitioned coarray signals are summed and averaged, the fourth-order auto-correlation tensor thereof is calculated; the position of the virtual sensor in the subarray .sub.(g.sub.1.sub.,g.sub.2.sub.) is expressed as:
.sub.(g.sub.1.sub.,g.sub.2.sub.)={(x,y)|X=p.sub.xd,y=p.sub.yd,−N.sub.x+g.sub.1≤p.sub.x≤−N.sub.x+g.sub.1+Y.sub.1−1,−N.sub.y+g.sub.2—p.sub.y≤−N.sub.y+g.sub.2+Y.sub.2−1}; the equivalent signals V.sub.(g1,g2) of the virtual subarray .sub.(g.sub.1.sub.,g.sub.2.sub.) is obtained by selecting corresponding position elements in the coarray signals V through the subarray .sub.(g.sub.1.sub.,g.sub.2.sub.).

5. The high-resolution accurate two-dimensional direction-of-arrival estimation method based on coarray tensor spatial spectrum searching with coprime planar array according to claim 1, wherein in addition to the CANDECOMP/PARACFAC decomposition, extraction of the multi-dimensional features of the fourth-order auto-correlation coarray tensor in step (6) can also be realized by a high-order singular value decomposition, which is specifically expressed as:
=×.sub.1D.sub.x×.sub.2D.sub.y×.sub.3D*.sub.x×.sub.4D*.sub.y, where ×.sub.Q represents a modulo Q inner product of the tensor and the matrix along the Q.sup.th dimension; represents a kernel tensor containing high-order singular values, D.sub.x∈.sup.Y.sup.1.sup.×Y.sup.1, D.sub.y∈.sup.Y.sup.2.sup.×Y.sup.2, D*.sub.x∈.sup.Y.sup.1.sup.×Y.sup.1 and D*.sub.y∈.sup.Y.sup.2.sup.×Y.sup.2 represent singular matrixes corresponding to four dimensions of ; the first K columns and the last (Y.sub.1−K) columns of D.sub.x are separated into a signal subspace D.sub.xs∈.sup.Y.sup.1.sup.×K and a noise subspace D.sub.xn ∈.sup.Y.sup.1.sup.×(Y.sup.1.sup.−K); and the first K columns and the last (Y.sub.2−K) columns of D.sub.y are separated into a signal subspace D.sub.ys ∈.sup.Y.sup.2.sup.×K and a noise subspace D.sub.yn∈.sup.Y.sup.2.sup.×(Y.sup.2.sup.−K).

6. The high-resolution accurate two-dimensional direction-of-arrival estimation method based on coarray tensor spatial spectrum searching with coprime planar array according to claim 5, wherein construction of the tensor spatial spectrum described in step (7) can also be implemented using the noise subspace obtained based on the high-order singular value decomposition, which is expressed as .sub.HOSVD({tilde over (θ)}, {tilde over (φ)}):
.sub.HOSVD({tilde over (θ)},{tilde over (φ)})=∥F({tilde over (θ)},{tilde over (φ)})×.sub.1D.sub.xnD.sub.xn.sup.H×.sub.2D.sub.ynD.sub.yn.sup.H∥.sup.−2, where, (⋅).sup.H represents a conjugate transposition operation; after obtaining the spatial spectrum function .sub.HOSVD({tilde over (θ)}, {tilde over (φ)}), the two-dimensional direction-of-arrival estimation of the source can be obtained according to a two-dimensional spectral peak searching process.

7. The high-resolution accurate two-dimensional direction-of-arrival estimation method based on coarray tensor spatial spectrum searching with coprime planar array according to claim 1, wherein the specific steps of searching for two-dimensional spectrum peaks after obtaining the spatial spectrum function .sub.CP({tilde over (θ)}, {tilde over (φ)}) in step (7) are: using a° as a step length to gradually increase the value of (θ, φ), wherein a search starting point of the two-dimensional direction-of-arrival ({tilde over (θ)}, {tilde over (φ)}) is (−90°, 0°), and an end point is (90°, 180°); a spatial spectrum value of one .sub.CP({tilde over (θ)}, {tilde over (φ)}) can be correspondingly calculated for each ({tilde over (θ)}, {tilde over (φ)}), so that one spatial spectrum corresponding to ({tilde over (θ)}, {tilde over (φ)}), {tilde over (θ)}∈[−90°, 90°], {tilde over (φ)}∈[0°, 180° ] can be constructed; there are K peaks in the spatial spectrum, and the values of ({tilde over (θ)}, {tilde over (φ)}) corresponding to the K peaks are the two-dimensional direction-of-arrival estimation of the source.

Description

BRIEF DESCRIPTION OF DRAWINGS

(1) FIG. 1 is an overall flow diagram of the present disclosure.

(2) FIG. 2 is a schematic structural diagram of a coprime planar array according to the present disclosure.

(3) FIG. 3 is a schematic diagram of an augmented virtual array structure derived from the present disclosure.

(4) FIG. 4 is a schematic diagram of a tensor spatial spectrum constructed by the present disclosure.

DESCRIPTION OF EMBODIMENTS

(5) Hereinafter, the technical solution of the present disclosure will be further explained in detail with reference to the drawings.

(6) In order to solve the problems of signal multi-dimensional spatial structural information loss and limited spatial spectrum resolution and precision performance existing in the existing method, the present disclosure provides a high-resolution accurate two-dimensional direction-of-arrival estimation method based on coarray tensor spatial spectrum searching with coprime planar array. Through statistical analysis of tensor signals received by the coprime planar array, coarray signals with spatial structure information of the virtual array are constructed; based on the multi-dimensional feature analysis of tensor statistics of virtual domain, the relationship between a virtual domain model and a tensor spatial spectrum is established, so as to realize a high-resolution accurate two-dimensional direction-of-arrival estimation method based on coarray tensor spatial spectrum searching. Referring to FIG. 1, the implementation steps of the present disclosure are as follows:

(7) Step 1: a coprime planar array is constructed. A receiving end constructs a coprime planar array using 4M.sub.xM.sub.y+N.sub.xN.sub.y−1 physical antenna array elements. As shown in FIG. 2: a pair of sparse uniform planar subarrays .sub.1 and .sub.2 are constructed on a planar coordinate system xoy, where .sub.1 contains 2M.sub.x×2M.sub.y antenna array elements, an array element spacing in the x-axis direction and the y-axis direction are N.sub.xd and N.sub.yd respectively, the coordinate of which on xoy is {(N.sub.xdm.sub.x, N.sub.ydm.sub.y), m.sub.x=0, 1, . . . , 2M.sub.x−1, m.sub.y=0, 1, . . . , 2M.sub.y−1}; .sub.2 contains N.sub.x×N.sub.y antenna array elements, an array element spacing in the x-axis direction and the y-axis direction are M.sub.xd and M.sub.yd respectively, and the coordinate of which on xoy is {(M.sub.xdn.sub.x, M.sub.ydn.sub.y), n.sub.x=0, 1, . . . , N.sub.x−1, n.sub.y=0, 1, . . . , N.sub.y−1}; here, M.sub.x, N.sub.x and M.sub.y, N.sub.y are respectively a pair of coprime integers, and M.sub.x<N.sub.x, M.sub.y<N.sub.y; the unit interval d is taken as half of the wavelength λ of the incident narrowband signal, that is, d=λ/2; .sub.1 and .sub.2 are subjected to subarray combination in a way of overlapping array elements at a (0,0) coordinate to obtain a coprime area array actually containing 4M.sub.xM.sub.y+N.sub.xN.sub.y−1 physical antenna array elements.

(8) Step 2: tensor signal modeling for the coprime planar array; assuming that there are K far-field narrowband incoherent signal sources from directions of {(θ.sub.1, φ.sub.1), (θ.sub.2, φ.sub.2), . . . , (θ.sub.K, φ.sub.K)}, the received signals of the sparse subarray .sub.1 of the coprime planar array is expressed by using a three-dimensional tensor signal .sub.1∈.sup.2M.sup.x.sup.×2M.sup.y.sup.×L (L is a sampling snapshot number) as:

(9) $x_1=\underset{k=1}{\overset{K}{.Math.}}{a}_{Mx}\left({\theta }_k,{\phi }_k\right)\circ a_{\mathrm{My}}\left({\theta }_k,{\phi }_k\right)\circ s_k+{𝒩}_1,$

(10) where s.sub.k=[.sub.k,1, s.sub.k,2, . . . , s.sub.k,L].sup.T is a multi-snapshot sampling signal waveform corresponding to a k.sup.th incident information source, [⋅].sup.T represents transposition operation, ∘ represents a vector outer product, .sub.1 is a noise tensor independent of each signal source, a.sub.Mx(θ.sub.k, φ.sub.k) and a.sub.My(θ.sub.k, φ.sub.k) are respectively steering vectors of .sub.1 in x-axis and y-axis directions, corresponding to the k.sup.th source with a direction-of-arrival of (θ.sub.k, φ.sub.k), which are expressed as:

(11) $a_{\mathrm{Mx}}\left({\theta }_k,{\phi }_k\right)={\left[1,e^{-j\pi u_1^{\left(2\right)}\sin \left({\phi }_k\right)\cos \left({\theta }_k\right)},\mathrm{.Math.},e^{-j\pi u_1^{\left(2M_x\right)}\sin \left({\phi }_k\right)\cos \left({\theta }_k\right)}\right]}^T,a_{\mathrm{My}}\left({\theta }_k,{\phi }_k\right)={\left[1,e^{-j\pi v_1^{\left(2\right)}\sin \left({\phi }_k\right)\cos \left({\theta }_k\right)},\mathrm{.Math.},e^{-j\pi v_1^{\left(2M_y\right)}\sin \left({\phi }_k\right)\cos \left({\theta }_k\right)}\right]}^T,$

(12) where u.sub.1.sup.(i.sup.1.sup.) (i.sub.1=1, 2, . . . , 2M.sub.x) and .sub.1.sup.(i.sup.2.sup.) (i.sub.2=1, 2, . . . , 2M.sub.y) respectively represent actual positions of the i.sub.1.sup.th physical antenna sensor and the i.sub.2.sup.th physical antenna sensor of the sparse subarray .sub.1 in the x-axis and y-axis directions, and u.sub.1.sup.(1)=0, .sub.1.sup.(1)=0, j=√{square root over (−1)}.

(13) Similarly, the received signals of the sparse subarray .sub.2 can be expressed by using another three-dimensional tensor .sub.2∈.sup.N.sup.x.sup.×N.sup.y.sup.×L as:

(14) $x_2=\underset{k=1}{\overset{K}{.Math.}}{a}_{\mathrm{Nx}}\left({\theta }_{k\prime }{\phi }_k\right)\circ a_{\mathrm{Ny}}\left({\theta }_{k\prime }{\phi }_k\right)\circ s_k+N_2,$

(15) where .sub.2 is a noise tensor independent of each signal source, a.sub.Nx(θ.sub.k, φ.sub.k) and a.sub.Ny(θ.sub.k, φ.sub.k) are respectively steering vectors of the sparse subarray .sub.2 in the X-axis and Y-axis directions, corresponding to the k.sup.th source with a direction-of-arrival of (θ.sub.k, φ.sub.k) which are expressed as:

(16) $a_{\mathrm{Nx}}\left({\theta }_k,{\phi }_k\right)={\left[1,e^{-j\pi u_2^{\left(2\right)}\sin \left({\phi }_k\right)\cos \left({\theta }_k\right)},\mathrm{.Math.},e^{-j\pi u_2^{\left(N_x\right)}\sin \left({\phi }_k\right)\cos \left({\theta }_k\right)}\right]}^T,a_{\mathrm{Ny}}\left({\theta }_k,{\phi }_k\right)={\left[1,e^{-j\pi v_2^{\left(2\right)}\sin \left({\phi }_k\right)\cos \left({\theta }_k\right)},\mathrm{.Math.},e^{-j\pi v_2^{\left(N_y\right)}\sin \left({\phi }_k\right)\cos \left({\theta }_k\right)}\right]}^T,$

(17) where u.sub.2.sup.(i.sup.3.sup.) (i.sub.3=1, 2, . . . , N.sub.x) and .sub.2.sup.(i.sup.4.sup.) (i.sub.4=1, 2, . . . , N.sub.y) respectively represent actual positions of a i.sub.3.sup.th physical antenna array element and a i.sub.4.sup.th physical antenna array element of the sparse subarray .sub.2 in the x-axis and y-axis directions, and u.sub.2.sup.(1)=0, .sub.2.sup.(1)=0;

(18) a second-order cross-correlation tensor ∈.sup.2M.sup.x.sup.×2M.sup.y.sup.×N.sup.x.sup.×N.sup.y of the received tensor signals .sub.1 and .sub.2 of the subarrays .sub.1 and .sub.2 is calculated as follows:

(19) $\hat{R}=\frac{1}{L}\underset{l=1}{\stackrel{L}{.Math.}}{x}_1\left(l\right)\circ x_2^{*}\left(l\right),$

(20) where, .sub.1(l) and .sub.2(l) respectively represent a l.sup.th slice of .sub.1 and .sub.2 in a third dimension (i.e., a snapshot dimension), and (⋅)* represents a conjugate operation.

(21) Step 3: the coarray signals based on a second-order cross-correlation tensor for the coprime planar array are derived. The second-order cross-correlation tensor of the tensor signal received by two subarrays of the coprime planar array can be ideally (a noise-free scene) modeled as:
=Σ.sub.k=1.sup.Kσ.sub.k.sup.2a.sub.Mx(θ.sub.k,φ.sub.k)∘a.sub.My(θ.sub.k,φ.sub.k)∘a*.sub.Nx(θ.sub.k,φ.sub.k)∘a*.sub.Ny(θ.sub.k,φ.sub.k),

(22) where σ.sub.k.sup.2 represents the power of the k.sup.th incident signal source; here, a.sub.Mx(θ.sub.k, φ.sub.k)∘a*.sub.Nx(θ.sub.k, φ.sub.k) in is equivalent to an augmented virtual domain along the x-axis, a.sub.My(θ.sub.k, φ.sub.k)∘a*.sub.Ny(θ.sub.k, φ.sub.k) is equivalent to an augmented virtual domain along the y-axis, and thereby the non-uniform virtual array can be obtained.

(23) As shown in FIG. 3, the position of each virtual sensor is expressed as:
={(−M.sub.xn.sub.xd+N.sub.xm.sub.xd,−M.sub.yn.sub.yd+N.sub.ym.sub.yd)|0≤n.sub.x≤N.sub.x−1,0≤m.sub.x≤2M.sub.x−1,0≤n.sub.y≤N.sub.y−1,0≤m.sub.y≤2M.sub.y−1}.

(24) In order to obtain the equivalent received signal corresponding to the augmented virtual domain area array , the first and third dimensions representing the spatial information in the x-axis direction in the cross-correlation tensor are combined into one dimension, and the second and fourth dimensions representing the spatial information in the y-axis direction are combined into one dimension. The dimensional combination of the tensor can be realized by PARAFAC-based tensor unfolding. Dimension sets .sub.1={1, 3} and ={2, 4} are defined, and then an ideal value (a noise-free scene) of the cross-correlation tensor is subjected to modulo {} PARAFAC-based tensor unfolding to obtain an ideal expression of the equivalent received signals V∈.sup.2M.sup.x.sup.N.sup.x.sup.×2M.sup.y.sup.N.sup.y of the augmented virtual array as:
V=Σ.sub.k=1.sup.Kσ.sub.k.sup.1a.sub.x(θ.sub.k,φ.sub.k)∘a.sub.y(θ.sub.k,φ.sub.k),

(25) where a.sub.x(θ.sub.k, φ.sub.k)=a*.sub.Nx(θ.sub.k, φ.sub.k).Math.a.sub.Mx(θ.sub.k, φ.sub.k) and a.sub.y(θ.sub.k, φ.sub.k)=a*.sub.Ny(θ.sub.k, φ.sub.k).Math.a.sub.My(θ.sub.k, φ.sub.k) are steering vectors of the augmented virtual array in the x-axis and y-axis directions, corresponding to k.sup.th signal source with a direction-of-arrival of (θ.sub.k, φ.sub.k); σ.sub.k.sup.2 represents the power of a k.sup.th incident signal source; .Math. represents a Kronecker product.

(26) Step 4: the coarray signals of a virtual uniform array is constructed. contains a virtual array with an x-axis distribution from (−N.sub.x+1)d to (M.sub.xN.sub.x+M.sub.x−1)d and a y-axis distribution from (−N.sub.y+1)d to (M.sub.yN.sub.y+M.sub.y−1)d; wherein there are D.sub.x×D.sub.y virtual sensors in total in , where D.sub.x=M.sub.xN.sub.x+M.sub.x+N.sub.x−1, D.sub.y=m.sub.yN.sub.y+M.sub.y+N.sub.y−1, and is expressed as:
={(x,y)|x=p.sub.xd,y=p.sub.yd,−N.sub.x+1≤p.sub.x≤M.sub.xN.sub.x+M.sub.x−1,−N.sub.y+1≤p.sub.y≤M.sub.yN.sub.y+M.sub.y−1},

(27) by selecting elements in the coarray signal V corresponding to the position of each virtual sensor of , the equivalent signals V∈.sup.D.sup.x.sup.×D.sup.y of the virtual array is obtained, and is expressed as:

(28) 0$\overline{V}=\underset{k=1}{\overset{K}{.Math.}}{\sigma }_k^2{b}_x\left({\theta }_k,{\phi }_k\right)\circ b_y\left({\theta }_k,{\phi }_k\right),$

(29) where b.sub.x(θ.sub.k, φ.sub.k)=[e.sup.−jπ(−N.sup.x.sup.+1)sin(φ.sup.k.sup.)cos(θ.sup.k.sup.), e.sup.−jπ(−N.sup.x.sup.+2)sin(φ.sup.k.sup.)cos(θ.sup.k.sup.), . . . , e.sup.−jπ(M.sup.x.sup.N.sup.x.sup.+M.sup.x.sup.−1)sin(φ.sup.k.sup.)cos(θ.sup.k.sup.)] and b.sub.y(θ.sub.k, φ.sub.k)=[e.sup.−jπ(−N.sup.y.sup.+1)sin(φ.sup.k.sup.)sin(θ.sup.k.sup.), e.sup.−jπ(−N.sup.y.sup.+2)sin(φ.sup.k.sup.)sin(θ.sup.k.sup.), . . . , e.sup.−jπ(M.sup.y.sup.N.sup.y.sup.+M.sup.y.sup.−1)sin(φ.sup.k.sup.)sin(θ.sup.k.sup.)] are steering vectors of the virtual array in the x-axis and y-axis directions, corresponding to k.sup.th source with a direction-of-arrival of (θ.sub.k, φ.sub.k).

(30) Step 5: a spatially smoothed auto-correlation coarray tensor is derived. The equivalent signal V of the virtual array is obtained from the above step. The equivalent signal V of the virtual described in step (5) saves spatial structure information of the virtual domain area array, however, since V can be regarded as a virtual domain signal of a single snapshot, the statistics thereof often have a rank deficiency problem; therefore, based on the idea of two-dimensional spatial smoothing, the virtual domain signal V is processed to construct multiple equivalent snapshot virtual domain subarray signals; after the virtual domain subarray signals are summed and averaged, the fourth-order self-correlation tensor thereof is calculated. The specific process is as below: in the virtual array , taking a subarray with a size of Y.sub.1×Y.sub.2 for every other array element along the x-axis and y-axis directions respectively to divide the virtual into L.sub.1×L.sub.2 uniform subarrays partly overlapping with each other; L.sub.1, L.sub.2, Y.sub.1, Y.sub.2 satisfy the following relationship:
Y.sub.1+L.sub.1−1=M.sub.xN.sub.x+M.sub.x+N.sub.x−1,
Y.sub.2+L.sub.2−1=M.sub.yN.sub.y+M.sub.y+N.sub.y−1;

(31) the above subarray is expressed as .sub.(g.sub.1.sub.,g.sub.2.sub.), g.sub.1=1, 2, . . . , L.sub.1, g.sub.2=1, 2, . . . , L.sub.2, and the position of the array element in .sub.(g.sub.1.sub.,g.sub.2.sub.) is expressed as:
.sub.(g.sub.1.sub.,g.sub.2.sub.)={(x,y)|X=p.sub.xd,y=p.sub.yd,−N.sub.x+g.sub.1≤p.sub.x≤−N.sub.x+g.sub.1+Y.sub.1−1,−N.sub.y+g.sub.2—p.sub.y≤−N.sub.y+g.sub.2+Y.sub.2−1};

(32) an equivalent signal V.sub.(g.sub.1.sub.,g.sub.2.sub.)∈.sup.Y.sup.1.sup.×Y.sup.2 of the virtual subarray .sub.(g.sub.1.sub.,g.sub.2.sub.) is obtained according to respective position elements in the coarray signals V corresponding to the subarray .sub.(g.sub.1.sub.,g.sub.2.sub.):
V.sub.(g.sub.1.sub.,g.sub.2.sub.)=Σ.sub.k=1.sup.Kσ.sub.k.sup.2(c.sub.x(θ.sub.k,φ.sub.k)e.sup.(g.sup.1.sup.−1)sin(φ.sup.k.sup.)cos(θ.sup.k.sup.))∘(c.sub.y(θ.sub.k,φ.sub.k)e.sup.(g.sup.2.sup.−1)sin(φ.sup.k.sup.)sin(θ.sup.k.sup.)),
where
c.sub.x(θ.sub.k,φ.sub.k)=[e.sup.−jπ(−N.sup.x.sup.+1)sin(φ.sup.k.sup.)cos(θ.sup.k.sup.),e.sup.−jπ(−N.sup.x.sup.+2)sin(φ.sup.k.sup.)cos(θ.sup.k.sup.), . . . ,e.sup.−jπ(M.sup.x.sup.N.sup.x.sup.+M.sup.x.sup.−1)sin(φ.sup.k.sup.)cos(θ.sup.k.sup.)] and
c.sub.y(θ.sub.k,φ.sub.k)=[e.sup.−jπ(−N.sup.y.sup.+1)sin(φ.sup.k.sup.)sin(θ.sup.k.sup.),e.sup.−jπ(−N.sup.y.sup.+2)sin(φ.sup.k.sup.)sin(θ.sup.k.sup.), . . . ,e.sup.−jπ(M.sup.y.sup.N.sup.y.sup.+M.sup.y.sup.−1)sin(φ.sup.k.sup.)sin(θ.sup.k.sup.)]
are steering vectors of the virtual subarray .sub.(1,1) in the x-axis and y-axis corresponding to the direction of (θ.sub.k, φ.sub.k); after the above operation, a total of L.sub.1×L.sub.2 coarray signals V.sub.(g.sub.1.sub.,g.sub.2.sub.) the dimensions of which are all Y.sub.1×Y.sub.2 are obtained; an average of the L.sub.1×L.sub.2 coarray signals VV.sub.(g.sub.1.sub.,g.sub.2.sub.) is taken to obtain a spatially smoothed signal {tilde over (V)}∈.sup.Y.sup.1.sup.×Y.sup.2:

(33) $\tilde{V}=\frac{1}{L_1{L}_2}\underset{p=1}{\overset{L_1}{.Math.}}\underset{q=1}{\overset{L_2}{.Math.}}{\overline{V}}_{\left(p,q\right)},$

(34) a fourth-order self-correlation tensor ∈.sup.Y.sup.1.sup.×Y.sup.2.sup.×Y.sup.1.sup.×Y.sup.2 for the coarray signals {tilde over (V)} is calculated as follows:
={tilde over (V)}∘{tilde over (V)}*.

(35) Step 6: signal and noise subspace classification is realized on the basis of multi-dimensional feature extraction of the auto-correlation coarray tensor. In order to construct a tensor spatial spectrum based on the idea of subspace classification, CANDECOMP/PARACFAC decomposition is performed on the fourth-order self-correlation tensor to extract multi-dimensional features, a result of which is expressed as follows:
=Σ.sub.k=1.sup.K{tilde over (c)}.sub.x(θ.sub.k,φk)∘{tilde over (c)}.sub.y(θ.sub.k,φ.sub.k)∘{tilde over (c)}*.sub.x(θ.sub.k,φ.sub.k)∘{tilde over (c)}*.sub.y(θ.sub.k,φ.sub.k),

(36) where {tilde over (c)}.sub.x(θ.sub.k, φ.sub.k)(k=1, 2, . . . , K) and {tilde over (c)}.sub.y(θ.sub.k, φ.sub.k)(k=1, 2, . . . , K) are two orthogonal factor vectors obtained by the CANDECOMP/PARACFAC decomposition, respectively representing spatial information in the x-axis and y-axis directions; C.sub.x=[{tilde over (c)}.sub.x(θ.sub.1, φ.sub.1), {tilde over (c)}.sub.x(θ.sub.2, φ.sub.2), . . . , {tilde over (c)}.sub.x(θ.sub.K, φ.sub.K)] and C.sub.y=[{tilde over (c)}.sub.y(θ.sub.1, φ.sub.1), {tilde over (c)}.sub.y(θ.sub.2, φ.sub.2), . . . , {tilde over (c)}.sub.y(θ.sub.K, φ.sub.K)] are factor matrixes; a space expanded by {{tilde over (c)}.sub.x(θ.sub.k, φ.sub.k)∘{tilde over (c)}.sub.y(θ.sub.k, φ.sub.k), k=1, 2 . . . , K} is taken and is recorded as span{{tilde over (c)}.sub.x(θ.sub.k, φ.sub.k)∘{tilde over (c)}.sub.y(θ.sub.k, φ.sub.k), k=1, 2, . . . , K} as a signal subspace; the signal subspace is expressed using a tensor .sub.s∈.sup.Y.sup.1.sup.×Y.sup.2.sup.×K where .sub.s(k) represents a k.sup.th slice of .sub.s along the third dimension, which is expressed as:
.sub.s(k)={tilde over (c)}.sub.x(θ.sub.k,φ.sub.k)∘{tilde over (c)}.sub.y(θ.sub.k,φ.sub.k);

(37) in order to obtain the noise subspace, the orthocomplements of the factor matrices C.sub.x and C.sub.y are calculated; the orthocomplement of C.sub.x is recorded as span{{tilde over (d)}.sub.x,h, h=1, 2, . . . , min(Y.sub.1, Y.sub.2)−K}, and the orthocomplement of C.sub.y is recorded as span{{tilde over (d)}.sub.y,h, h=1, 2, . . . , min(Y.sub.1, Y.sub.2) K}, where min(∘) represents the operation of taking the minimum value; then span{{tilde over (d)}.sub.x,h∘{tilde over (d)}.sub.y,h, h=1, 2, . . . , min(Y.sub.1, Y.sub.2)−K} is taken as the noise subspace, and the tensor .sub.n∈.sup.Y.sup.1.sup.×Y.sup.2.sup.×(min(Y.sup.1.sup.,Y.sup.2.sup.)−K) is used to express the noise subspace, and .sub.n(h) represents a h.sup.th slice of .sub.n along the third dimension, which is expressed as:
.sub.n(h)={tilde over (d)}.sub.x,h∘{tilde over (d)}.sub.y,h;

(38) in addition to the CANDECOMP/PARACFAC decomposition, the extraction of the multi-dimensional features of the fourth-order self-correlation tensor in the virtual domain in step (6) can also be realized by a high-order singular value decomposition, which is specifically expressed as:
=×.sub.1D.sub.x×.sub.2D.sub.y×.sub.3D*.sub.x×.sub.4D*.sub.y,

(39) where ×.sub.Q represents a modulo Q inner product of the tensor and the matrix along the Q.sup.th dimension; represents a kernel tensor containing high-order singular values, D.sub.x∈.sup.Y.sup.1.sup.×Y.sup.1, D.sub.y∈.sup.Y.sup.2.sup.×Y.sup.2, D*.sub.x ∈.sup.Y.sup.1.sup.×Y.sup.1 and D*.sub.y∈.sup.Y.sup.2.sup.×Y.sup.2 represent singular matrixes corresponding to four dimensions of ; the first K columns and the last Y.sub.1−K columns of D.sub.x are separated into a signal subspace D.sub.xs∈.sup.Y.sup.1.sup.×K and a noise subspace D.sub.xn ∈.sup.Y.sup.1.sup.×(Y.sup.1.sup.−K); the first K columns and the last Y.sub.2−K columns of D.sub.y are separated into a signal subspace D.sub.ys∈.sup.Y.sup.2.sup.×K and a noise subspace D.sub.yn∈.sup.Y.sup.2.sup.×(Y.sup.2.sup.−K).

(40) Step 7: high-resolution accurate two-dimensional direction-of-arrival estimation based on coarray tensor spatial spectrum searching. A two-dimensional direction-of-arrival ({tilde over (θ)}, {tilde over (φ)}) for spectral peak searching is defined, a {tilde over (θ)}∈[−90°, 90°], {tilde over (φ)}∈[0°, 180° ], and steering information F({tilde over (θ)}, {tilde over (φ)})∈.sup.Y.sup.1.sup.×Y.sup.2 corresponding to the virtual array are constructed, which is expressed as:
F({tilde over (θ)},{tilde over (φ)})=c.sub.x({tilde over (θ)},{tilde over (φ)})∘c.sub.y({tilde over (θ)},{tilde over (φ)}),

(41) a tensor spatial spectrum function .sub.CP({tilde over (θ)}, {tilde over (φ)}) is constructed using the noise subspace obtained from the CANDECOMP/PARACFAC decomposition, which is expressed as follows:
.sub.CP({tilde over (θ)},{tilde over (φ)})=∥<Z.sub.n×.sub.{1,2}F({tilde over (θ)},{tilde over (φ)})>∥.sub.F.sup.−2,

(42) where <.sub.{Q}> represents a modulo {Q} contraction operation of two tensors along a Q.sup.th dimension, which requires a same size of the Q.sup.th dimension of the two tensors; ∥⋅∥.sub.F represents a Frobenius norm; z.sub.n∈.sup.Y.sup.1.sup.×Y.sup.2.sup.×(min(Y.sup.1.sup.,Y.sup.2.sup.)−K) and F({tilde over (θ)}, {tilde over (φ)})∈.sup.Y.sup.1.sup.×Y.sup.2 are subjected to modulo {1, 2} reshaping along the 1.sup.st and 2.sup.nd dimensions to obtain one vector p∈.sup.min(Y.sup.1.sup.,Y.sup.2.sup.)−K;

(43) after obtaining the spatial spectrum function .sub.CP({tilde over (θ)}, {tilde over (φ)}), the estimation of two-dimensional direction-of-arrival of the incident source is obtained by searching for the two-dimensional spectral peak. The specific steps are: using a° as a step length to gradually increase the value of (θ, φ), wherein a search starting point of the two-dimensional direction-of-arrival ({tilde over (θ)}, {tilde over (φ)}) is (−90°, 0°), and an end point is (90°, 180°); a spatial spectrum value of one .sub.CP({tilde over (θ)}, {tilde over (φ)}) can be correspondingly calculated for each ({tilde over (θ)}, {tilde over (φ)}), so that one spatial spectrum corresponding to ({tilde over (θ)}, {tilde over (φ)}), {tilde over (θ)}∈[−90°, 90°], {tilde over (φ)}∈[0°, 180° ] can be constructed; there are K peaks in the spatial spectrum, and the values of ({tilde over (θ)}, {tilde over (φ)}) corresponding to the K peaks are the two-dimensional direction-of-arrival estimation of the information source.

(44) The effect of the present disclosure will be further described with a simulation example.

(45) The construction of the tensor spatial spectrum can be implemented using the noise subspace obtained based on the high-order singular value decomposition, which is expressed as .sub.HOSVD({tilde over (θ)}, {tilde over (φ)}):
.sub.HOSVD({tilde over (θ)},{tilde over (φ)})=∥F({tilde over (θ)},{tilde over (φ)})×.sub.1D.sub.xnD.sub.xn.sup.H×.sub.2D.sub.ynD.sub.yn.sup.H∥.sup.−2,

(46) where, (⋅).sup.H represents a conjugate transposition operation; after obtaining the spatial spectrum function .sub.HOSVD(θ, φ), the two-dimensional direction-of-arrival estimation of the source can be obtained according to a two-dimensional spectral peak searching process.

(47) The effects of the present disclosure will be further described in the following in combination with examples of simulation.

(48) Example of Simulation:

(49) A coprime array is used to receive an incident signal, and its parameters are selected as M.sub.x=2, M.sub.y=2, N.sub.x=3, N.sub.y=3, that is, the coprime array of the architecture consists of 4M.sub.xM.sub.y+N.sub.xN.sub.y−1=24 physical elements. It is assumed that the number of incident narrowband signals is 1, and the azimuth angle and elevation angle of the incident direction are [45°, 50° respectively. L=500 sampling snapshots and 10 dB input signal-to-noise ratio are used for simulation experiment.

(50) The spatial spectrum of the high-resolution accurate two-dimensional direction-of-arrival estimation method based on coarray tensor spatial spectrum searching of a coprime planar array is shown in FIG. 4. It can be seen that the proposed method can effectively construct a two-dimensional spatial spectrum, in which there is a sharp spectral peak in the two-dimensional direction-of-arrival corresponding to the incident source, and the values of X axis and Y axis corresponding to the spectral peak are the elevation angle and azimuth angle of the incident source.

(51) To sum up, the invention fully considers the multi-dimensional structural information of coprime planar array signals, constructs coarray signals with spatial structural information of a virtual planar array by using tensor signal modeling, establishes a subspace classification idea based on auto-correlation coarray tensor multi-dimensional feature extraction by analyzing tensor statistical characteristics, builds the connection between a coprime planar array virtual domain model and a tensor spatial spectrum, and solves the signal mismatch problem of the coprime planar array. Meanwhile, by using two tensor feature extraction method, namely tensor decomposition and high-order singular value decomposition, the construction mechanism of high-precision and high-resolution tensor spatial spectrum is proposed, and compared with the existing method, a breakthrough is made in spatial spectrum resolution and performance of the two-dimensional direction-of-arrival estimation accuracy.

(52) The above is only the preferred embodiment of the present disclosure. Although the present disclosure has been disclosed as a preferred embodiment, it is not intended to limit the present disclosure. Without departing from the scope of the technical solution of the present disclosure, any person familiar with the field can make many possible changes and modifications to the technical solution of the present disclosure by using the methods and technical contents disclosed above, or modify them into equivalent embodiments with equivalent changes. Therefore, any simple modifications, equivalent changes and modifications made to the above embodiments according to the technical essence of the present disclosure are still within the scope of protection of the technical solution of the present disclosure.