CN110196417A - The bistatic MIMO radar angle estimating method concentrated based on emitted energy - Google Patents

Abstract

The invention discloses a bistatic MIMO radar angle estimation method based on transmission energy concentration, which is realized as follows: in the case of a known target relative to the transmission array angle region, the transmission beam domain matrix is designed and scaled by solving a convex optimization problem; Define the transmit waveform based on the transmit beam domain matrix to obtain the received echo model; perform matched filtering on the echo model, and convert the filtered data into column vectors to construct a new echo model; solve the newly constructed echo model Its autocorrelation matrix, and decompose the eigenvalues of the autocorrelation matrix to obtain the signal subspace containing the wave departure angle and arrival angle information; obtain the estimated value of the target wave departure angle and arrival angle according to the signal subspace; establish a mapping relationship to Compensate the wave departure angle difference error in the process of solving the convex optimization problem, and obtain the optimal estimation value of the target wave departure angle. The invention solves the problem that the transmission energy is not concentrated in the prior art, improves the angle measurement accuracy, and can be used for target detection.

Description

An Angle Estimation Method for Bistatic MIMO Radar Based on Transmit Energy Concentration

technical field

The invention belongs to the technical field of radar, and further relates to a bistatic MIMO radar angle estimation method, which can be used to improve the problem of energy loss caused by uniform transmission gains in all directions of the traditional bistatic MIMO radar in the angle measurement process.

Background technique

Compared with the traditional phased array radar, MIMO radar has attracted extensive attention in recent years. MIMO radar is based on the conventional phased array radar and requires the transmitting antenna to transmit incompletely correlated waveforms. When transmitting orthogonal waveforms, it can be used. The coverage of the signal in the airspace is wider, and during signal processing, multiple beams can be formed to cover the airspace at the same time.

In a bistatic MIMO radar, the wave departure angle DOD refers to the angle between the direction of the radar's transmitted signal and the normal of the transmitting antenna, and the angle of arrival DOA refers to the angle between the direction of the target echo and the normal to the receiving antenna. The angle measurement of wave departure and arrival angles is an important research area for bistatic MIMO radars. At present, the angle measurement methods of bistatic MIMO radar wave departure angle and arrival angle are mainly based on subspace class algorithms, and ESPRIT algorithm and MUSIC algorithm are commonly used in subspace class algorithms. Compared with the MUSIC algorithm, the ESPRIT algorithm uses the rotation-invariant technology to avoid the spectral peak search of the MUSIC algorithm through two sub-arrays with identical characteristics, and has better robustness.

In the document "Angle estimation using ESPRIT in MIMO radar", a bistatic MIMO radar angle estimation method based on ESPRIT algorithm is proposed. In the above document, the ESPRIT algorithm adopts waveform diversity technology, that is, its transmit waveform is completely orthogonal. The echo data solves the autocorrelation matrix and performs eigenvalue decomposition to obtain the signal subspace containing the information of the departure angle and the arrival angle of the target wave; however, due to the use of completely orthogonal transmit waveforms, the transmit gain in all directions is relatively uniform , where the angular region of the target relative to the transmit array is known, it actually reduces the coherence gain of the transmit and results in degraded goniometric performance, which is a significant disadvantage of traditional bistatic MIMO radar goniometric.

SUMMARY OF THE INVENTION

The purpose of the present invention is to propose a bistatic MIMO radar angle estimation method based on the concentration of transmission energy to improve the angle measurement accuracy of the departure angle and the arrival angle of the target wave.

The technical solution of the present invention is that, when the angular area of the target relative to the transmitting array is known, the transmit beam domain matrix is designed and scaled by solving a convex optimization problem, the transmit waveform based on the transmit beam domain matrix is defined, and then the transmit beam domain matrix is used. The received echo model is obtained from the waveform, matched filtering is performed on the echo model, and the filtered data is converted into a column vector to construct a new echo model, the autocorrelation matrix is solved for the newly constructed echo model, and the autocorrelation matrix is Perform eigenvalue decomposition to obtain the signal subspace containing the wave departure angle and arrival angle information, and obtain the estimated value of the target wave departure angle and arrival angle according to the obtained signal subspace, and finally establish a mapping relationship. In the process of solving the convex optimization problem The difference error of the wave departure angle caused by the error is compensated, so as to obtain the optimal estimation value of the target wave departure angle. Its implementation steps include the following:

(1) Set the number of transmitting array elements of the bistatic MIMO radar to M, the number of receiving array elements to be N, the total transmit energy of the bistatic MIMO radar to be E, the spatial dimension of the transmit beam domain to be L, and the total amount of the radar in a coherent processing interval CPI. The number of echo pulses is Q, and there are G targets in the target airspace;

(2) According to the angular area of the target relative to the transmit array, the transmit beam domain matrix W is designed using the convex optimization method, and W is scaled to satisfy tr(W ^* W ^T )=L, and the scaled transmit beam is obtained Domain matrix W', where [ ] ^* represents the conjugate of the matrix, [ ] ^T represents the transpose of the matrix, and tr( ) represents the trace of the matrix;

(3) According to the transmit waveform based on transmit beam domain Calculate the echo signal matrix X _q of the qth received pulse, and then take the q value for X _q from 1 to Q, and obtain the echo signal matrix X=[X ₁ , X ₂ ,...,X _q ,... ,X _Q ],

where [ ] ^* denotes the conjugate of the matrix, Represents L mutually orthogonal waveforms transmitted by the ^bistatic MIMO radar transmitting array, and SSH = _IL , _IL represents the L×L identity matrix, t=1,2,...,T,T represents each the number of samples in the pulse period, represents the waveform transmitted by the mth array element in the transmit waveform S, q=1,...,Q represents the pulse number;

(4) matched filtering is performed on the echo signal X _q of the qth received pulse, and the matched filtered data matrix Y _q is converted into a column vector y _q , and then a new echo signal matrix Y is constructed;

(5) calculate the autocorrelation matrix R of the newly constructed echo signal matrix Y, and carry out eigenvalue decomposition to the autocorrelation matrix R to determine the signal subspace E _s ;

(6) According to the signal subspace E _s , calculate the estimated value of the departure angle of the target wave and the estimated angle of arrival Among them, k=1,...,G, represents the target number;

(7) Establish a noise-free echo signal matrix model:

where A is the transmit steering matrix, B is the receive steering matrix, C is the reflection coefficient matrix, [ ] ^H represents the conjugate transpose of the matrix, and ⊙ represents the Khatri-Rao product;

(8) According to the noise-free echo matrix model in (7), when the array structure of the transmitting antenna and the receiving antenna is a half-wavelength uniformly spaced linear array ULA, the following mapping relationship can be obtained:

where J _t,1 =[I _L-1 ,0], J _t,2 =[0,I _L-1 ], I _L-1 is the L-1 order unit matrix, θ _i represents the target relative to the transmitting array The angle value of the ith sampling point in the angle area, a(θ _i ) is the launch steering vector of the target airspace, represents the mapping angle value of the ith sampling point in the angular region of the target relative to the emission array, One-to-one correspondence with θ _i , i=1,...,I, where I is the number of sampling points in the angular region of the target relative to the transmitting array;

(9) According to the mapping relationship obtained in (8), calculate the wave departure angle corresponding to the signal model in (7)

(10) The wave departure angle corresponding to the noise-free signal model Find the wave departure angle estimate from each target in The closest wave departure angle Then find the corresponding θ _i through the mapping relationship in (8), and use it as the optimal estimation value of the corresponding target wave departure angle.

Compared with the prior art, the present invention has the following advantages:

First, the present invention uses the convex optimization method to design the transmit beam domain matrix of the MIMO radar, so that the transmit energy of the MIMO radar is concentrated in the space sector where the target is located, thereby reducing the loss of transmit energy.

Second, compared with the ESPRIT angle measurement method of the traditional MIMO radar, the present invention enhances the signal-to-noise ratio of the target echo signal and improves the wave departure angle due to the high transmission energy of the transmitting array in the angular region of the space sector where the target is located. and the angular accuracy of the angle of arrival.

Third, the present invention performs error compensation on the estimated value of the target wave off-angle by establishing a mapping relationship, thereby reducing the wave off-angle error caused by using the convex optimization method to solve the transmit beam domain matrix.

Description of drawings

Fig. 1 is the realization flow chart of the present invention;

2 is a comparison diagram of the transmission energy distribution of the bistatic MIMO radar based on the concentrated transmission energy in the present invention and the transmission energy distribution of the traditional bistatic MIMO radar;

Fig. 3 is the comparison diagram of the root mean square error of the present invention and the traditional angle measurement method to the angle measurement of two unrelated targets with signal-to-noise ratio;

FIG. 4 is a comparison diagram of the root mean square error of the present invention and the traditional angle measurement method for angle measurement of two uncorrelated targets as a function of the number of pulses.

Detailed ways

The embodiments and effects of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, but not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

1, the implementation steps of this embodiment are as follows:

Step 1, set bistatic MIMO radar parameters

Set the number of transmitting array elements of the bistatic MIMO radar to M, M>0, the number of receiving array elements to be N, N>0, the total transmit energy of the bistatic MIMO radar to be E>0, and the spatial dimension of the transmit beam domain to be L, L <M, the total number of echo pulses in one coherent processing interval CPI of the radar is q, q=1, 2,...Q, Q is the total number of echo pulses in one coherent processing interval CPI of the radar, in the target airspace The number of targets is G, G>0;

Step 2, according to the position of the target relative to the angular region of the transmitting array, use the convex optimization method to design the transmitting beam domain matrix W.

The convex optimization method refers to the method of solving the optimization problem in which the objective function is a convex function and the set to which the variable belongs is a convex set.

In this example, the convex optimization method is used to design the transmit beam domain matrix W according to the angular area of the target relative to the transmit array. The implementation is as follows:

(2.1) Determine the angle area Θ of the target relative to the launch array and its complement angle area

The angular area Θ of the target relative to the transmitting array is set as [δ _min ,δ _max ], where δ _min represents the minimum expected angle at which the target appears, and δ _max represents the maximum expected angle at which the target appears;

Complement angular region of the target relative to the angular region Θ of the transmit array are [-90°, δ _min -1°] and [δ _max +1°, 90°];

(2.2) According to the parameters in (2.1), set a virtual launch steering vector of L×1 pointing to the angle area Θ When the array structure of the transmitting antenna and the receiving antenna is a half-wavelength uniformly spaced linear array ULA:

Among them, [ ] ^T represents the transpose of the vector, θ _i ∈Θ, i=1,2,...,I, I represents the total number of sampling points of Θ;

(2.3) According to the parameters in (2.1), when the array structure of the transmitting antenna and the receiving antenna is a half-wavelength uniformly spaced linear array ULA, the transmission steering vector a (θ _i ) of the pointing angle area Θ and the pointing angle area The launch steering vector a(θ _β ) of , respectively, is expressed as:

Among them, [·] ^T represents the transpose of the vector, θ _i ∈Θ, i=1,2,...,I, I represents the total number of sampling points of Θ, β=1,2,...,J, J means The total number of sampling points;

(2.4) According to the parameters of (2.1), (2.2) and (2.3), design the transmit beam domain matrix W by solving the following convex optimization problem:

Among them, [·] ^H represents the conjugate transpose of the matrix; ||·|| represents the 2-norm of the matrix; α represents the maximum acceptable W ^H a(θ _i ) and The error between the two is required to be as small as possible on the premise of ensuring that the above optimization problem has a solution, and the empirical value is generally 0<α<1.

Step 3: The transmit beam domain matrix W is scaled to satisfy tr(W ^* W ^T )=L, and the scaled transmit beam domain matrix W' is obtained.

Step 4: Obtain the echo signal matrix X according to the scaled transmit beam domain matrix W'.

(4.1) According to the scaled transmit beam domain matrix, the transmit waveform based on transmit beam domain is obtained:

where [ ] ^* denotes the conjugate of the matrix, represents L mutually orthogonal waveforms, and ^SSH = _IL , _IL represents the L×L unit matrix, t=1,2,...,T,T represents the number of samples per pulse period, represents the mth quadrature waveform in the transmitted waveform S;

(4.2) Determine the transmit steering matrix A and the receive steering matrix B:

When the transmitting and receiving arrays are half-wavelength equally spaced linear arrays ULA, the departure angle and arrival angle of the k-th target are denoted by θ _k and Representation, k=1,2,...,G, then the transmit steering matrix A and the receive steering matrix B can be expressed as:

A=[a(θ ₁ ),a(θ ₂ ),...a(θ _k ),...,a(θ _G )]

in, represents the launch steering vector of the kth target,

Represents the received steering vector of the k-th target;

(4.3) The reflection coefficient diagonal matrix Λ _q of the G targets at the qth echo pulse is expressed as:

Λ _q =diag(c _q ),

Among them, diag(·) represents the diagonalization operation, c _q =[α _1,q ,α _2,q ,…α _k,q ,…,α _G,q ] ^T represents the qth echo of the G target The reflection coefficient vector of the pulse, α _k,q represents the reflection coefficient of the kth target in the qth echo pulse;

(4.4) Calculate the echo signal matrix X of the qth received pulse according to the transmit waveform in (4.1), the transmit steering matrix A, the receive steering matrix B in (4.2), and the reflection coefficient diagonal matrix Λ _q in (4.3). _q :

Among them, [ ] ^T represents the matrix transpose, [ ] ^* represents the matrix conjugate; N _q represents the noise matrix of the qth echo pulse, whose mean value is zero and obeys the Gaussian white noise distribution;

(4.5) Take the value of q for X _q from 1 to Q in turn, and obtain the echo signal matrix X=[X ₁ , X ₂ ,...,X _q ,...,X _Q ].

In step 5, a new echo signal matrix Y is constructed by using the echo signal X _q of the qth received pulse obtained in step 4 .

(5.1) Multiply the signal matrix X _q of the q-th echo pulse to the right by ^SH and perform matched filtering to obtain the signal matrix Y _q of the q-th pulse after matched filtering:

Among them, [ ] ^T represents the matrix transpose, [ ] ^H represents the matrix conjugate transpose; S is the L mutually orthogonal waveforms transmitted by the bistatic MIMO radar transmitting array; represents the noise matrix of the qth pulse after matched filtering, Λ _q represents the diagonal matrix of reflection coefficients of G targets at the qth echo pulse; A is the transmit steering matrix; B is the receive steering matrix; represents the scaled transmit beam domain matrix;

(5.2) According to the formula in (5.1), arrange Y _q in columns to obtain the signal column vector y _q of the qth pulse after matched filtering:

in, Represents the column vector formed by the matched filtered noise, vec( ) represents the quantization of the matrix vector; c _q represents the reflection coefficient vector of the qth echo pulse of the G targets; ⊙ represents the Khatri-Rao product;

(5.3) Take the value of q from 1 to Q, and according to the formula in (5.2), the following matrices are obtained:

The signal column vector y ₁ of the first pulse after matched filtering to the signal column vector y _Q of the Q-th pulse after matched filtering is denoted as a new echo signal matrix: Y=[y ₁ , y ₂ , .. .,y _q ,...,y _Q ];

Denote the reflection coefficient vector c ₁ of the G targets at the first pulse to the reflection coefficient vector c _Q of the G targets at the Q-th pulse as a reflection coefficient matrix: C=[c ₁ ,c ₂ ,..., c _q ,...,c _Q ];

The noise matrix of the first pulse after matching filtering Noise matrix to the Qth pulse after matched filtering Denoted as the noise matrix after matched filtering:

(5.4) Substitute the formula in (5.2) into Y=[y ₁ , y ₂ ,...,y _q ,...,y _Q ], according to the reflection coefficient matrix C in (5.3) and the matched filtered noise matrix Get the new echo signal matrix Y:

Step 6, according to the new echo signal matrix Y obtained in step 5, determine the signal subspace _Es .

(6.1) Calculate the autocorrelation matrix R of the new echo signal matrix Y:

Among them, [ ] ^H represents the conjugate transpose of the matrix, and y _q is the signal column vector of the qth pulse after matched filtering.

(6.2) Perform eigenvalue decomposition on the autocorrelation matrix R to get:

Among them, uk represents the k-th eigenvalue of the autocorrelation matrix R, v _k represents the eigenvector corresponding to the _k -th eigenvalue _uk of the autocorrelation matrix R, σ ² represents the noise power, and I _LN represents the LN order unit matrix;

(6.3) Determine the signal subspace

The eigenvector v ₁ corresponding to the first eigenvalue u ₁ of the autocorrelation matrix R in (6.2) to the eigenvector v _G corresponding to the G-th eigenvalue u _G of the autocorrelation matrix R form a MN×G dimension matrix by column , denote this matrix as the signal subspace E _s .

Step 7, according to the signal subspace E _s , calculate and obtain the estimated value of the departure angle of the target wave and the estimated angle of arrival

(7.1) Define the transmit angle matrix Ψ _t and the receive angle matrix Ψ _r

When the transmitting and receiving arrays are half-wavelength equally spaced linear arrays ULA, the departure angle and arrival angle of the k-th target are denoted by θ _k and Representation, k=1,2,...,G, defines that the transmit angle matrix Ψ _t and the receive angle matrix Ψ _r have the following structure:

Ψ _t =T ^-1 Φ _t T

Ψ _r =T ^-1 Φ _r T

Among them, Φ _t is the eigenvalue matrix of Ψ _t ,

Φ _r is the eigenvalue matrix of Ψ _r ,

T is a specific G×G non-singular matrix;

(7.2) According to the properties of the signal subspace and the structure of the radar array, the following two equations are obtained:

Among them, the symbol represents the Kronecker product, Es represents the signal subspace, J _{t ,1} =[I _L-1 ,0], J _t,2 =[0,I _L-1 ], J _r,1 =[ _IN-1 , 0], J _r,2 = [0, I _N-1 ] represents the selection matrix, I _L-1 represents the identity matrix of (L-1)×(L-1), and I _N-1 represents (N-1) ×(N-1) unit matrix, I _N represents the N×N unit matrix, and _IL represents the L×L unit matrix;

(7.3) Substitute (7.1) into the equation in (7.2) to solve the transmit angle matrix Ψ _t and the receive angle matrix Ψ _r :

in, Represents the pseudo-inverse of the matrix;

(7.4) Perform eigendecomposition on the obtained transmit angle matrix Ψ _t and receive angle matrix Ψ _r :

Among them, γ _k represents the k-th eigenvalue of the transmission angle matrix Ψ _t , q _k represents the eigenvector corresponding to the k-th eigenvalue γ _k of the transmission angle matrix Ψ _t , and μ _k represents the k-th eigenvalue of the receiving angle matrix Ψ _r eigenvalue, f _k represents the eigenvector corresponding to the k-th eigenvalue μ _k of the receiving angle matrix Ψ _r ;

(7.5) According to γ _k and μ _k in (7.4), in this embodiment, it is considered that γ _k and μ _k contain the angle information of the same target, and the specific angle pairing method does not belong to the innovative content of the present invention, and the following formula is used Obtain the estimated value of the wave departure angle of the kth target and the estimated angle of arrival of the k-th target

Among them, angle( ) represents the operation of taking the phase;

Step 8, establish a noise-free echo signal matrix model:

where A is the transmit steering matrix, B is the receive steering matrix, C is the reflection coefficient matrix, [ ] ^H represents the conjugate transpose of the matrix, and ⊙ represents the Khatri-Rao product.

Step 9, according to the noise-free echo signal matrix model in step 8, establish a mapping relationship, and calculate the wave departure angle corresponding to the signal model

(9.1) According to the noise-free echo matrix model in step 8, when the array structure of the transmitting antenna and the receiving antenna is a half-wavelength uniformly spaced linear array ULA, the following mapping relationship is obtained:

Among them, J _t,1 =[I _L-1 ,0], J _t,2 =[0,I _L-1 ], I _L-1 is an L-1 order identity matrix, θ _i ∈Θ,i=1 ,2,...,I, I represents the total number of sampling points of Θ, a(θ _i ) is the launch steering vector of the target airspace, represents the mapping angle value of θ _i , One-to-one correspondence with θ _i ;

(9.2) According to the mapping relationship obtained in (9.1), calculate I wave departure angles corresponding to the signal model in step 8

Step 10: Obtain the optimal estimated value of the off-angle of each target wave by searching.

(10.1) The wave departure angle calculated in (9.2) In i=1,...,I, find the estimated value of the wave departure angle from each target in step 7 The closest wave departure angle

(10.2) According to the mapping relationship in (9.1), find the wave departure angle from (10.1) The corresponding sampling angle θ _i , and the sampling angle θ _i is used as the optimal estimation value of the corresponding target wave departure angle.

The effects of the present invention will be further described through the following simulation experiments.

1. Simulation experimental conditions:

Assume that the number of transmitting antennas of the bistatic MIMO radar is M=8, the number of receiving antennas is N=4, the transmitting array and the receiving array are half-wavelength equally spaced line arrays, the total transmitting energy E=8, the angular area of the target relative to the transmitting array Θ=[-15°, 15°], it is assumed that there are two irrelevant targets in Θ, which are located at and The number of Monte Carlo experiments was P=500.

2. Simulation content

Simulation 1, under the above conditions, set the transmit beam domain dimension L=6 in the present invention, the maximum acceptable W ^H a(θ _i ) and The error α=0.04; the energy distribution of the bistatic MIMO radar based on the transmission energy concentration in the present invention and the energy distribution of the traditional bistatic MIMO radar are simulated respectively, and the results are shown in FIG. 2 .

As can be seen from Figure 2, the transmission energy of the traditional bistatic MIMO radar is uniform, but the transmission energy of the bistatic MIMO radar of the present invention is concentrated in the angular area of the target relative to the transmission array, the loss of transmission energy is smaller, and the target echo signal The signal-to-noise ratio is higher.

Simulation 2, under the above-mentioned simulation conditions, set the total echo pulse number Q=50 in a coherent processing interval CPI of the radar, and in the present invention, set the transmission beam domain dimension L=6, set the maximum acceptable W in Θ ^H a(θ _i ) and The error α=0.04; respectively simulate the present invention and three traditional angle measurement methods, the traditional ESPRIT method, the traditional PM method and the traditional U-ESPRIT method, the root mean square error RMSE changes with the signal-to-noise ratio, the results As shown in Figure 3.

It can be seen from FIG. 3 that the root mean square error RMSE of the bistatic MIMO radar angle measurement of the present invention is smaller than that of other angle measurement methods, and the angle measurement accuracy is higher.

Simulation 3, under the above-mentioned simulation conditions, set the signal-to-noise ratio SNR=0db of the radar, and in the present invention, set the transmission beam domain dimension L=6, in Θ, the maximum acceptable W ^H a (θ _i ) and The error between them is α=0.04; simulating the present invention and three traditional angle measurement methods, the traditional ESPRIT method, the traditional PM method and the traditional U-ESPRIT method, the root mean square error RMSE varies with the total number of echo pulses , the results are shown in Figure 4.

It can be seen from FIG. 4 that the root mean square error RMSE of the bistatic MIMO radar angle measurement of the present invention is smaller than that of other angle measurement methods, and the angle measurement accuracy is higher.

The root mean square error is defined as:

in, and represents the optimal estimation of the wave departure angle and arrival angle of the pth Monte Carlo test of the kth target, θ _k and Represents the true value of the wave departure angle and arrival angle of the kth target.

The above are only specific embodiments of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed by the present invention. Included within the scope of protection of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

Claims (7)

1. a kind of bistatic MIMO radar angle estimation method (based on the concentrated bistatic MIMO radar angle measurement optimization method of transmission energy) based on transmission energy, it is characterized in that, comprise as follows: (1) Set the number of transmitting array elements of the bistatic MIMO radar to M, the number of receiving array elements to be N, the total transmit energy of the bistatic MIMO radar to be E, the spatial dimension of the transmit beam domain to be L, and the total amount of the radar in a coherent processing interval CPI. The number of echo pulses is Q, and there are G targets in the target airspace; (2) Use the convex optimization method to design the transmit beam domain matrix W according to the angular area of the target relative to the transmit array, and scale W to satisfy tr(W ^* W ^T )=L, and obtain the scaled transmit beam domain Matrix W', where [ ] ^* represents the conjugate of the matrix, [ ] ^T represents the transpose of the matrix, and tr( ) represents the trace of the matrix; (3) According to the transmit waveform based on transmit beam domain Calculate the echo signal matrix X _q of the qth received pulse, and then take q values from 1 to Q for X _q to obtain the echo signal matrix X=[X ₁ , X ₂ ,...,X _q ,. ..,X _Q ], where [ ] ^* denotes the conjugate of the matrix, Represents L mutually orthogonal waveforms transmitted by the ^bistatic MIMO radar transmitting array, and SSH = _IL , _IL represents the L×L identity matrix, t=1,2,...,T,T represents each the number of samples in the pulse period, represents the waveform transmitted by the mth array element in the transmit waveform S, q=1,...,Q represents the pulse number; (4) matched filtering is performed on the echo signal X _q of the qth received pulse, and the matched filtered data matrix Y _q is converted into a column vector y _q , and then a new echo signal matrix Y is constructed; (5) calculate the autocorrelation matrix R of the newly constructed echo signal matrix Y, and carry out eigenvalue decomposition to the autocorrelation matrix R to determine the signal subspace E _s ; (6) According to the signal subspace E _s , calculate the estimated value of the departure angle of the target wave and the estimated angle of arrival Among them, k=1,...,G, represents the target number; (7) Establish a noise-free echo signal matrix model: where A is the transmit steering matrix, B is the receive steering matrix, C is the reflection coefficient matrix, [ ] ^H represents the conjugate transpose of the matrix, and ⊙ represents the Khatri-Rao product; (8) According to the noise-free echo matrix model in (7), when the array structure of the transmitting antenna and the receiving antenna is a half-wavelength uniformly spaced linear array ULA, the following mapping relationship can be obtained: where J _t,1 =[I _L-1 ,0], J _t,2 =[0,I _L-1 ], I _L-1 is the L-1 order unit matrix, θ _i represents the target relative to the transmitting array The angle value of the ith sampling point in the angle area, a(θ _i ) is the launch steering vector of the target airspace, represents the mapping angle value of the ith sampling point in the angular region of the target relative to the emission array, One-to-one correspondence with θ _i , i=1,...,I, I is the number of sampling points in the angular area of the target relative to the emission array; (9) According to the mapping relationship obtained in (8), calculate the wave departure angle corresponding to the signal model in (7) (10) The wave departure angle corresponding to the noise-free signal model Find the wave departure angle estimate from each target in The closest wave departure angle Then find the corresponding θ _i through the mapping relationship in (8), and use it as the optimal estimation value of the corresponding target wave departure angle. 2. method as claimed in claim 1 is characterized in that, utilizes convex optimization method to design transmit beam domain matrix W according to the position of space sector where target is located in (2), realizes as follows: (2a) When the array structure of the transmitting antenna and the receiving antenna is a half-wavelength uniformly spaced linear array ULA, an L×1 virtual transmitting steering vector can be Expressed as: Among them, [ ] ^T represents the transpose of the vector, and θ represents the wave departure angle of the target; (2b) According to the parameters designed in (2a), the transmit beam domain matrix W is obtained by solving the following convex optimization problem: where [·] ^H represents the conjugate transpose of the matrix, ||·|| represents the 2-norm of the matrix, θ _i ∈Θ, i=1,2,...,I, Θ represents the target relative to the emission The angular area of the array, I represents the total number of sampling points of Θ, represents the complement angle region of Θ, and J represents The total number of sampling points, [ ] ^T represents the transpose of the matrix, α represents the maximum acceptable W ^H a(θ _i ) and error between. 3. The method according to claim 1, wherein the echo signal matrix X=[X ₁ , X ₂ ,..., X _q ,..., X _Q ] is obtained in (3), which realizes as follows: (3a) When the transmitting and receiving arrays are half-wavelength uniformly spaced linear arrays ULA, the wave departure angle and arrival angle of the k-th target are denoted by θ _k and Representation, k=1,2,...,G, then the transmit steering matrix A and the receive steering matrix B can be expressed as: A=[a(θ ₁ ),a(θ ₂ ),...a(θ _k ),...,a(θ _G )] in, represents the launch steering vector of the kth target, represents the received steering vector of the k-th target; (3b) The reflection coefficient diagonal matrix Λq of G targets at the _qth echo pulse can be expressed as: Λ _q =diag(c _q ) Among them, diag(·) represents the diagonalization operation, and c _q =[α _1,q ,α _2,q ,...α _k,q ,...,α _G,q ] ^T indicates that the G target is in the first The reflection coefficient vector of the q echo pulses, α _k,q represents the reflection coefficient of the k th target in the q th echo pulse; (3c) According to the transmission waveform and the parameters in (3a) and (3b), calculate the echo signal matrix X _q of the qth received pulse: Among them, [ ] ^T represents the matrix transpose, [ ] ^* represents the matrix conjugate; N _q represents the noise matrix of the qth echo pulse, whose mean value is zero and obeys the Gaussian white noise distribution; (3d) Take the value of q for X _q from 1 to Q in sequence, and obtain the echo signal matrix X=[X ₁ , X ₂ ,...,X _q ,...,X _Q ]. 4. method as claimed in claim 1 is characterized in that, the new echo signal matrix Y of building in (4), realizes as follows: (4a) Multiply the signal matrix X _q of the q-th echo pulse to the right by ^SH and perform matched filtering to obtain the signal matrix Y _q of the q-th pulse after matched filtering: Among them, [ ] ^T represents the matrix transpose, [ ] ^H represents the matrix conjugate transpose; S is the L mutually orthogonal waveforms transmitted by the bistatic MIMO radar transmitting array; represents the noise matrix of the qth pulse after matched filtering, Λ _q represents the diagonal matrix of reflection coefficients of G targets at the qth echo pulse; A is the transmit steering matrix; B is the receive steering matrix; (4b) According to the formula in (4a), arrange Y _q in columns to obtain the signal column vector y _q of the qth pulse after matched filtering: in, Represents the column vector formed by the matched filtered noise, vec( ) represents the quantization of the matrix vector; c _q represents the reflection coefficient vector of the qth echo pulse of the G targets; ⊙ represents the Khatri-Rao product; (4c) Take the value of q from 1 to Q. According to the formula in (4b), the following matrices are obtained respectively: The signal column vector y ₁ of the first pulse after matched filtering to the signal column vector y _Q of the Q-th pulse after matched filtering is denoted as a new echo signal matrix: Y=[y ₁ , y ₂ ,... ,y _q ,...,y _Q ]; The reflection coefficient vector c ₁ of G targets in the first pulse to the reflection coefficient vector c _Q of G targets in the Qth pulse, denoted as reflection coefficient matrix: C=[c ₁ ,c ₂ ,...,c _q ,...,c _Q ]; Noise matrix of the first pulse after matched filtering Noise matrix to the Qth pulse after matched filtering Denoted as the noise matrix after matched filtering: (4d) Substitute the formula in (4b) into Y=[y ₁ , y ₂ ,...,y _q ,...,y _Q ], and obtain a new echo signal matrix according to the parameters in (4c) Y is: 5. method as claimed in claim 1 is characterized in that, the autocorrelation matrix R in (5), its calculation is as follows: Among them, [ ] ^H represents the conjugate transpose of the matrix, and y _q is the signal column vector of the qth pulse after matched filtering. 6. The method of claim 1, wherein the signal subspace _Es in (5) is determined as follows: (5a) Perform eigenvalue decomposition on the autocorrelation matrix R to get: Among them, uk represents the k-th eigenvalue of the autocorrelation matrix R, v _k represents the eigenvector corresponding to the _k -th eigenvalue _uk of the autocorrelation matrix R, σ ² represents the noise power, and I _LN represents the LN order unit matrix; (5b) From the eigenvector v1 corresponding to the first eigenvalue u1 of the autocorrelation matrix R in (5a) to the eigenvector _vG corresponding to the _Gth eigenvalue _{uG of} the covariance matrix R, MN× G-dimensional matrix, denote this matrix as the signal subspace E _s . 7. method as claimed in claim 1 is characterized in that, in (6), calculate target wave off-angle estimation value and the estimated angle of arrival The implementation is as follows: (6a) When the transmitting and receiving arrays are half-wavelength uniformly spaced linear arrays ULA, the wave departure angle and arrival angle of the kth target are calculated by θ _k and Representation, k=1,2,...,G, then the transmitting angle matrix Ψ _t and the receiving angle matrix Ψ _r can be expressed as: Ψ _t =T ^-1 Φ _t T Ψ _r =T ^-1 Φ _r T Among them, Φ _t is the eigenvalue matrix of Ψ _t , Φ _r is the eigenvalue matrix of Ψ _r , T is a specific G×G non-singular matrix; (6b) According to the properties of the signal subspace and the structure of the radar array, the following two equations are obtained: Among them, the symbol represents the Kronecker product, ES represents the signal subspace, J _t,1 =[I _{L -1} ,0], J _t,2 =[0,I _L-1 ], J _r,1 =[ _IN-1 , 0], J _r,2 = [0, I _N-1 ] represents the selection matrix, I _L-1 represents the identity matrix of (L-1)×(L-1), and I _N-1 represents (N-1) ×(N-1) unit matrix, I _N represents the N×N unit matrix, and _IL represents the L×L unit matrix; (6c) Substitute (6a) into the equation in (6b) to solve the transmit angle matrix Ψ _t and the receive angle matrix Ψ _r : in, Represents the pseudo-inverse of the matrix; (6d) Eigendecomposition of the transmit angle matrix Ψ _t and the receive angle matrix Ψ _r Among them, γ _k represents the k-th eigenvalue of the transmission angle matrix Ψ _t , q _k represents the eigenvector corresponding to the k-th eigenvalue γ _k of the transmission angle matrix Ψ _t , and μ _k represents the k-th eigenvalue of the receiving angle matrix Ψ _r eigenvalue, f _k represents the eigenvector corresponding to the k-th eigenvalue μ _k of the receiving angle matrix Ψ _r ; (6e) According to the parameters in (6d), use the following formula to obtain the estimated value of the wave departure angle of the kth target and the estimated angle of arrival of the k-th target Among them, angle(·) represents the operation of taking the phase.