Multi-target motion parameter estimation method and device based on MIMO radar

Abstract

The application provides a multi-target motion parameter estimation method and device based on MIMO radar, which comprises: performing down-conversion processing on echo signals observed by each receiving array of a MIMO radar system to obtain identification results of baseband signals corresponding to each receiving array; constructing a covariance matrix corresponding to each receiving array based on the identification results; performing eigenvalue decomposition on each covariance matrix, and determining a noise subspace corresponding to each receiving array based on the results of the eigenvalue decomposition of each receiving array; and determining an estimation result of a motion parameter of each target to be measured based on the noise subspace corresponding to each receiving array and an array response matrix; wherein the array response matrix comprises array response characteristics of the receiving array and signal characteristics of the echo signals observed by the receiving array. The application can effectively reduce the calculation load in a multi-target scene, thereby overcoming the limitation of direct estimation methods in the application of multi-target scenes.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of data processing, and in particular to a multi-target motion parameter estimation method and device based on MIMO radar. BACKGROUND

[0002] With the continuous breakthrough of radar technology, Multiple-Input Multiple-Output (MIMO) radar has attracted extensive attention from researchers. Benefiting from the excellent characteristics of signal waveform, frequency, polarization diversity and spatial diversity of radar, MIMO radar has made great progress in military, civil and other aspects in recent years.

[0003] As a high-precision target motion parameter estimation method, the direct estimation method can approach the Cramer-Rao Lower Bound (CRLB) under lower signal-to-noise ratio conditions. Compared with the traditional two-step estimation method, the direct estimation method does not need to estimate the intermediate parameters in the implementation process, so it does not need to match and associate the parameters in the multi-target scene. Therefore, the direct estimation method has been widely used in active and passive estimation fields.

[0004] However, when the existing direct estimation method estimates the motion parameters of each target in the multi-target scene, the computational load will linearly increase with the increase of the number of targets, which greatly limits the practicability of the existing direct estimation method in MIMO radar. SUMMARY

[0005] The present application provides a multi-target motion parameter estimation method and device based on MIMO radar to solve the defect of high computational load in the multi-target scene in the prior art.

[0006] The present application provides a multi-target motion parameter estimation method based on MIMO radar, comprising:

[0007] Downconvert the echo signals observed by each receiving array of the MIMO radar system to obtain the identification results of the baseband signals corresponding to each receiving array; wherein the echo signals observed by each receiving array are formed by superimposing the signals reflected by multiple targets to be tested to the receiving array;

[0008] Based on the identification results of the baseband signals corresponding to each receiving array, construct the covariance matrix corresponding to each receiving array respectively;

[0009] Perform eigenvalue decomposition on the covariance matrix corresponding to each receiving array respectively, and determine the noise subspace corresponding to each receiving array based on the results of the eigenvalue decomposition corresponding to each receiving array respectively.

[0010] determine an estimation result of the motion parameter of each of the to-be-tested targets based on the noise subspace corresponding to each of the receiving arrays and an array response matrix; wherein the array response matrix comprises an array response characteristic of the receiving array and a signal characteristic of the echo signal observed by the receiving array.

[0011] According to the MIMO radar-based multi-target motion parameter estimation method provided in the application, the identification result of the baseband signal corresponding to each of the receiving arrays is used to construct a covariance matrix corresponding to each of the receiving arrays, comprising:

[0012] performing signal discretization processing on the identification result of the baseband signal corresponding to the receiving array to obtain a baseband signal vector corresponding to the receiving array; wherein the baseband signal vector comprises an expected signal matrix and a noise signal matrix;

[0013] performing subspace expansion on the expected signal matrix to obtain an expected signal expansion matrix;

[0014] updating the baseband signal vector based on the expected signal expansion matrix, and constructing the covariance matrix corresponding to the receiving array based on the updated baseband signal vector.

[0015] According to the MIMO radar-based multi-target motion parameter estimation method provided in the application, the expected signal matrix comprises a steering vector item, a Doppler frequency offset item, a baseband signal item, a time delay item and a scattering coefficient item;

[0016] the subspace expansion on the expected signal matrix to obtain an expected signal expansion matrix, comprising:

[0017] determining an array response matrix corresponding to the expected signal matrix based on the steering vector item, the Doppler frequency offset item and the baseband signal item, and determining a snapshot data vector corresponding to the expected signal matrix based on the time delay item and the scattering coefficient item;

[0018] determining the expected signal expansion matrix based on the array response matrix and the snapshot data vector; wherein the steering vector item, the Doppler frequency offset item and the time delay item are all constructed based on the motion parameter item of each of the to-be-tested targets.

[0019] According to the MIMO radar-based multi-target motion parameter estimation method provided in the application, the result of the eigenvalue decomposition corresponding to the receiving array comprises a plurality of eigenvalues and a characteristic vector corresponding to each of the eigenvalues; the noise subspace corresponding to each of the receiving arrays is determined based on the result of the eigenvalue decomposition corresponding to each of the receiving arrays, comprising:

[0020] ordering each of the eigenvalues corresponding to the receiving array, and determining a target vector from each of the eigenvectors corresponding to the receiving array according to the ordering result;

[0021] determining the noise subspace corresponding to the receiving array based on the target vector.

[0022] According to the application, the method for estimating the motion parameters of multiple targets based on MIMO radar is provided, and the estimation result of the motion parameters of each target is determined based on the noise subspace corresponding to each receiving array and the array response matrix, which includes:

[0023] The estimation result of the motion parameters of each target is determined based on the orthogonal characteristics of each column vector in the array response matrix corresponding to each receiving array and the noise subspace corresponding to the corresponding receiving array.

[0024] According to the application, the method for estimating the motion parameters of multiple targets based on MIMO radar is provided, and the estimation result of the motion parameters of each target is determined based on the orthogonal characteristics of each column vector in the array response matrix corresponding to each receiving array and the noise subspace corresponding to the corresponding receiving array, which includes:

[0025] A cost function is constructed based on each column vector in the array response matrix corresponding to each receiving array and the noise subspace corresponding to each receiving array, and the cost function includes a group of motion parameter variables.

[0026] The estimation result of the motion parameters of each target is determined based on the output value of the cost function in the preset value interval of the motion parameters of each target.

[0027] According to the application, the method for estimating the motion parameters of multiple targets based on MIMO radar is provided, and the cost function is constructed based on each column vector in the array response matrix corresponding to each receiving array and the noise subspace corresponding to each receiving array, which includes:

[0028] A column vector function model is constructed based on each column vector in the array response matrix corresponding to each receiving array and the motion parameter variable, and the column vector function model is a function model irrelevant to each target.

[0029] The cost function is constructed based on the column vector function model and the noise subspace corresponding to each receiving array.

[0030] The application further provides a multi-target motion parameter estimation device based on a MIMO radar, comprising: a first processing module, configured to perform down-conversion processing on echo signals observed by each receiving array of a MIMO radar system to obtain an identification result of a baseband signal corresponding to each receiving array; wherein the echo signals observed by each receiving array are formed by superposition of signals reflected by multiple to-be-detected targets to the receiving array;

[0031] a second processing module, configured to construct a covariance matrix corresponding to each receiving array based on the identification result of the baseband signal corresponding to each receiving array;

[0032] a third processing module, configured to perform eigenvalue decomposition on the covariance matrix corresponding to each receiving array respectively, and determine a noise subspace corresponding to each receiving array based on a result of the eigenvalue decomposition corresponding to each receiving array;

[0033] a fourth processing module, configured to determine an estimation result of a motion parameter of each to-be-detected target based on the noise subspace corresponding to each receiving array and an array response matrix; wherein the array response matrix comprises array response characteristics of the receiving array and signal characteristics of the echo signals observed by the receiving array.

[0034] The application further provides an electronic device, comprising a memory, a processor and a computer program stored in the memory and capable of running on the processor, wherein the processor implements the multi-target motion parameter estimation method based on the MIMO radar when executing the program.

[0035] The application further provides a non-transitory computer readable storage medium, which stores a computer program, wherein the computer program is executed by a processor to implement the multi-target motion parameter estimation method based on the MIMO radar.

[0036] The application provides a multi-target motion parameter estimation method and device based on a MIMO radar. BRIEF DESCRIPTION OF DRAWINGS

[0037] In order to more clearly illustrate the technical solutions in the application or the prior art, the drawings needed to be used in the embodiments or the prior art description will be briefly introduced. Obviously, the drawings in the following description are some embodiments of the application, and other drawings can be obtained by those skilled in the art without any creative effort on the basis of these drawings.

[0038] Figure 1 FIG. 1 is a flowchart of the multi-target motion parameter estimation method based on the MIMO radar provided by the application;

[0039] Figure 2 FIG. 2 is a simulation scene diagram of the motion parameter estimation of multiple targets based on the MIMO radar system provided by the application;

[0040] Figure 3 FIG. 3 is a comparison diagram of the estimation accuracy of the initial position of target 1 by the method provided by the application and the ML-DPD estimation method in a multi-target scene;

[0041] Figure 4 FIG. 4 is a comparison diagram of the estimation accuracy of the speed of target 1 by the method provided by the application and the ML-DPD estimation method in a multi-target scene;

[0042] Figure 5 FIG. 5 is a comparison diagram of the estimation accuracy of the initial position of target 2 by the method provided by the application and the ML-DPD estimation method in a multi-target scene;

[0043] Figure 6 is a comparison chart of the estimation accuracy of the speed of target 2 by the method of the present application and the ML-DPD estimation method in the multi-target scenario provided by the present application;

[0044] Figure 7 is a structural schematic diagram of the multi-target motion parameter estimation device based on MIMO radar provided by the present application;

[0045] Figure 8 is a structural schematic diagram of the electronic device provided by the present application. DETAILED DESCRIPTION

[0046] To make the objectives, technical solutions and advantages of the present application clearer, the technical solutions in the present application will be described clearly and completely below in conjunction with the drawings in the present application. Obviously, the described embodiments are some embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of protection of the present application.

[0047] The multi-target motion parameter estimation method based on MIMO radar provided by the present application will be described below in conjunction with Figures 1-6 The multi-target motion parameter estimation method based on MIMO radar provided by the present application is executed by electronic devices such as computers located in a data fusion processing center or hardware and / or software therein. As shown in Figure 1 The multi-target motion parameter estimation method based on MIMO radar provided by the present application at least includes the following steps:

[0048] S101, performing down-conversion processing on the echo signals observed by each receiving array of a MIMO radar system to obtain an identification result of the baseband signals corresponding to each receiving array; wherein the echo signals observed by each receiving array are formed by superimposing signals reflected to the receiving array by multiple to-be-measured targets;

[0049] S102, constructing a covariance matrix corresponding to each receiving array based on the identification result of the baseband signals corresponding to each receiving array;

[0050] S103, performing eigenvalue decomposition on the covariance matrix corresponding to each receiving array, and determining a noise subspace corresponding to each receiving array based on the result of the eigenvalue decomposition corresponding to each receiving array;

[0051] S104, determining an estimation result of the motion parameters of each to-be-measured target based on the noise subspace corresponding to each receiving array and an array response matrix; wherein the array response matrix includes array response characteristics of the receiving array and signal characteristics of the echo signals observed by the receiving array.

[0052] In this embodiment, the MIMO radar system includes M T transmit antennas and M R receive arrays, each of which is provided with L receive antennas with a spacing of d. For any transmit antenna, K pulse signals are continuously transmitted, and the pulse repetition period of the pulse signals is denoted as T r , the number of targets to be detected is P (P > 1), and it is assumed that the distance between the target to be detected and each receive array satisfies the far-field condition.

[0053] For any receive array of the MIMO radar system, the echo signal observed by the receive array in the pulse repetition period corresponding to the kth (1≤k≤K) pulse signal is formed by the signals reflected by P targets to be detected to the receive array from the M T transmit antennas. The echo signal observed by the receive array is down-converted to remove the carrier signal, and the identification result of the baseband signal observed by each receive array is obtained. For example, the identification result of the baseband signal can be as shown in equation (1):

[0054]

[0055] In equation (1), r n,k (t) is the identification result of the baseband signal corresponding to the nth (1≤n≤M r ) receive array at time t (t ∈ (0, T R )) in the pulse repetition period corresponding to the kth pulse signal; α n,p is the steering vector of the signal reflected by the pth (1≤p≤P) target to be detected relative to the nth receive array, which can be determined based on the angle of arrival in the echo signal; β n,m,k,p is the scattering coefficient of the kth pulse signal transmitted by the mth (1≤m≤M T ) transmit antenna and reflected by the pth target to be detected to the nth receive array; s m (t) is the initial baseband signal transmitted by the mth transmit antenna at time t; τ n,m,k,p is the propagation delay of the kth pulse signal transmitted by the mth transmit antenna and reflected by the pth target to be detected to the nth receive array, which can be determined based on the time of arrival in the echo signal; and τ n,m,p is the propagation delay of any pulse signal transmitted by the mth transmit antenna and reflected by the pth target to be detected to the nth receive array, where it can be considered that the time delay variation of the pulse echo in the observation time period of the continuous K pulse repetition periods is much smaller than the bandwidth B of the initial baseband signal, i.e., τ n,m,p = τ n,m,1,p .n,m,1,p a propagation time delay of a first pulse signal transmitted by the mthtransmitting antenna and reflected by the pthto-be-detected target and then propagating to the nthreceiving array; a Doppler frequency offset caused by the motion of the pthto-be-detected target relative to the mthtransmitting antenna and the nthreceiving array to each pulse signal corresponding echo signal, which can be determined based on the Doppler frequency offset in the echo signal; f0is the frequency of the carrier signal, wherein the frequencies of the carrier signals used by each transmitting antenna can be the same, all being f0; w n,k (t) is an additive white Gaussian noise signal at time t within the pulse repetition period corresponding to the kthpulse signal affecting the output signals of each receiving antenna on the nthreceiving array, wherein the additive white Gaussian noise signals on each receiving antenna are independent of each other and have the same power σ 2 .

[0056] wherein α n,p , τ n,m,k,p and may be represented based on the motion parameters of the corresponding to-be-detected target, the position information of each transmitting antenna of the MIMO radar system and the position information of the receiving antennas in the corresponding receiving array, for example, α n,p , τ n,m,k,p and may be respectively as shown in equations (2)-(4):

[0057] α n,p = [1, exp(-j2πf m dcosθ n,p / c), …, exp(-j2πf m (L-1)dcosθ n,p / c] T (2)

[0058]

[0059]

[0060] In the formula, f m is the frequency of the carrier signal transmitted by the mthtransmitting antenna, which can be f0; c is the speed of light; the superscript T represents transposition; p p is the initial position of the pthto-be-detected target, p p = [x p , y p ], x p and y p are the coordinate values of the pthto-be-detected target on the x-axis and y-axis at the initial time, respectively; is the position of the mthtransmitting antenna, and These are the coordinates of the m-th transmitting antenna on the x-axis and y-axis, respectively. This represents the position of the first receiving antenna (reference element) in the nth receiving array. and Let x and y be the coordinates of the first receiving antenna in the nth receiving array on the x-axis and y-axis, respectively, where l is the position of the l-th receiving antenna in the nth receiving array. 1≤l≤L;v p Let v be the velocity of the p-th target to be measured. p =[v xp ,v yp ], v xp and v yp These represent the velocities of the p-th target along the x-axis and y-axis, respectively; cosθ n,p Let cosθ be the incident angle of the pulse signal after reflection from the p-th target and onto the n-th receiving array. It can be understood that the angular change of the target relative to each receiving array during the observation period of K consecutive pulse repetition cycles can be ignored, i.e., cosθ n,p It remains constant over K consecutive pulse repetition cycles; cosθ n,p It can be shown in equation (5):

[0061]

[0062] After obtaining the identification results of the baseband signals corresponding to each receiving array, for any receiving array in the MIMO radar system, the covariance matrix corresponding to that receiving array can be constructed based on the identification results of the baseband signals of that receiving array, thereby obtaining the covariance matrix corresponding to each receiving array.

[0063] By performing eigenvalue decomposition on the covariance matrix corresponding to each receiving array, the signal subspace and noise subspace corresponding to each receiving array can be obtained. For any receiving array in the MIMO radar system, the array response matrix corresponding to that receiving array can include the array response characteristics of that receiving array, such as the steering vector of the signal reflected by each target relative to the receiving array; at the same time, the array response matrix corresponding to that receiving array can also include the signal characteristics of the echo signal observed by the receiving array, such as the characteristics of the signal propagating to the receiving array after the initial baseband signal transmitted by each transmitting antenna is reflected by each target, and the Doppler frequency offset caused by the motion of each target relative to each transmitting antenna and the receiving array to the echo signal corresponding to each pulse signal.

[0064] After obtaining the noise subspaces corresponding to the receiving arrays, the estimation results of the motion parameters of the targets can be determined based on the noise subspaces corresponding to the receiving arrays and the array response matrices corresponding to the receiving arrays. The motion parameters can include velocity and initial position. Since the array response characteristics of the receiving arrays to different targets are different, and the signal characteristics of the signals reflected by different targets are different, the motion parameters of the targets can be decoupled based on the noise subspaces corresponding to the receiving arrays and the array response matrices corresponding to the receiving arrays, so that the calculation load does not increase linearly with the number of targets in a multi-target scenario.

[0065] In the embodiment of the application, the echo signals observed by the receiving arrays of the MIMO radar system are down-converted to obtain the identification results of the baseband signals corresponding to the receiving arrays, and the covariance matrices corresponding to each receiving array are constructed based on the identification results of the baseband signals corresponding to the receiving arrays. The eigenvalue decomposition is performed on the covariance matrices corresponding to the receiving arrays, and the noise subspaces corresponding to the receiving arrays are determined based on the results of the eigenvalue decomposition. The estimation results of the motion parameters of the targets are determined based on the noise subspaces corresponding to the receiving arrays and the array response matrices. The array response matrix includes the array response characteristics of the corresponding receiving array and the signal characteristics of the echo signal. Therefore, the motion parameters of the targets can be decoupled based on the noise subspaces corresponding to the receiving arrays and the array response matrices corresponding to the receiving arrays. In a multi-target scenario, the calculation load does not increase linearly with the number of targets, which effectively reduces the calculation load and overcomes the limitations of the direct estimation method in a multi-target scenario.

[0066] In the exemplary embodiment, the covariance matrix corresponding to each receiving array is constructed based on the identification results of the baseband signals corresponding to the receiving arrays, including:

[0067] The identification results of the baseband signals corresponding to the receiving arrays are subjected to signal discretization processing to obtain a baseband signal vector corresponding to the receiving array. The baseband signal vector includes an expected signal matrix and a noise signal matrix.

[0068] The expected signal matrix is subjected to subspace expansion to obtain an expected signal expansion matrix.

[0069] The baseband signal vector is updated based on the expected signal expansion matrix, and the covariance matrix corresponding to the receiving array is constructed based on the updated baseband signal vector.

[0070] In the embodiment, for any receiving array of the MIMO radar system, the baseband signal vector corresponding to the receiving array comprises signal values of the recognition result of the baseband signal corresponding to the receiving array at different sampling time points. For example, the baseband signal vector r s (t) obtained after sampling at the sampling time point t = h / f n,k (h = 0,..., N-1) is of length N, where f s is the sampling frequency. n,k The baseband signal vector r

[0071] (t) can be expressed as formula (6):

[0072] r n,k = A n,k 1 + w n,k (6)

[0073] wherein, 1 represents an all-one column vector of size P x 1; w n,k represents a noise signal matrix of length NL; and A n,k represents an expected signal matrix of size NL x P, which is specifically expressed as formula (7):

[0074] A n,k = [r n,k,1 ,..., r n,k,P ] (7)

[0075] wherein, r n,k,p is a vector of length NL obtained after sampling the expected signal of the kth pulse signal reflected by the pth target to be detected to the nth receiving array.

[0076] The expected signal matrix A n,k may be subjected to subspace expansion to obtain an expected signal expansion matrix, and the expected signal expansion matrix is used to replace A n,k 1 in the baseband signal vector r n,k to obtain an updated baseband signal vector.

[0077] As an optional implementation, the expected signal matrix comprises a steering vector term, a Doppler frequency offset term, a baseband signal term, a time delay term and a scattering coefficient term.

[0078] The expected signal matrix is subjected to subspace expansion to obtain an expected signal expansion matrix, comprising:

[0079] The desired signal spreading matrix is determined based on the array response matrix and the snapshot data vector; wherein the steering vector term, the Doppler frequency offset term and the time delay term are all constructed based on the motion parameter term of each of the to-be-tested targets.

[0080] For example, r n,k,p may be shown as formula (8):

[0081] r n,k,p = Θ n,p F n,p S n,p D n,k,p β n,k,p (8)

[0082] In the formula, Θ n,p is a steering vector matrix between the pth to-be-tested target and the nth receiving array, I N is an N*N dimensional unit matrix, is a Kronecker product; F n,p is a Doppler frequency offset matrix between the pth to-be-tested target and the nth receiving array, F n,m,p is a simplified form of the matrix without specific meaning, which can be shown as formula (9) specifically, and diag(·) is a diagonal operation. n,p S is a baseband signal matrix between the pth to-be-tested target and the nth receiving array, s n,m,p is a simplified form of the matrix without specific meaning, which can be shown as formula (10) specifically, and blkdiag(·) represents a block diagonal operation; D n,k,p is a time delay matrix between the pth to-be-tested target and the nth receiving array within the pulse repetition period corresponding to the kth pulse signal, which can be shown as formula (11) specifically; β n,k,p is a scattering coefficient matrix between the pth to-be-tested target and the nth receiving array within the pulse repetition period corresponding to the kth pulse signal, which can be shown as formula (12) specifically.

[0083]

[0084] s n,m,p = [s m (t0-τ n,m,p ),…,s m (t N-1 -τ n,m,p ) T (10)

[0085]

[0086]

[0087] Thus, the baseband signal vector r in formula (6) n,k may be shown as formula (13):

[0088] r n,k = Θ n F n S n D n,k β n,k +w n,k (13)

[0089] In the formula, Θ n is the steering vector matrix corresponding to the nth receiving array, which can be used as a steering vector term, and can be shown as formula (14) specifically; F n is the Doppler frequency offset matrix corresponding to the nth receiving array, which can be used as a Doppler frequency offset term, and can be shown as formula (15) specifically; S n is the baseband signal matrix corresponding to the nth receiving array, which can be used as a baseband signal term, and can be shown as formula (16) specifically; D n,k is the time delay matrix corresponding to the nth receiving array within the pulse repetition period corresponding to the kth pulse signal, which can be used as a time delay term, and can be shown as formula (17) specifically; β n,k is the scattering coefficient matrix corresponding to the nth receiving array within the pulse repetition period corresponding to the kth pulse signal, which can be used as a scattering coefficient term, and can be shown as formula (18) specifically.

[0090] Θ n = [Θ n,1 ,…, Θ n,P ] (14)

[0091] F n = blkdiag[F n,1 ,…, F n,P ] (15)

[0092] S n = blkdiag[S n,1 ,…, S n,P ] (16)

[0093] D n,k = blkdiag[D n,k,1 ,…, D n,k,P ] (17)

[0094]

[0095] In implementations, the array response matrix corresponding to the desired signal matrix A can be determined based on the steering vector term, the Doppler frequency offset term, and the baseband signal term, and the snapshot data vector corresponding to the desired signal matrix A can be determined based on the time delay term and the scattering coefficient term. That is, the array response matrix includes both the array response characteristics of the corresponding receiving array and the signal characteristics of the echo signal observed by the corresponding receiving array.

[0096] For example, the desired signal matrix A n,k The corresponding array response matrix Q n = Θ n F n S n The desired signal matrix A n,k The corresponding snapshot data vector b n,k = D n,k β n,k The desired signal spread matrix can be expressed as Q n b n,k . Wherein the snapshot data vector includes the observation signals at each sampling time within the pulse repetition period corresponding to the kth pulse signal. Thus, the baseband signal vector r n,k in equation (6) can be expressed as shown in equation (19):

[0097] r n,k = Q n b n,k +w n,k (19).

[0098] After obtaining the updated baseband signal vector, a covariance matrix corresponding to the receiving array can be constructed based on the updated baseband signal vector. For example, the covariance matrix corresponding to the nth receiving array can be expressed as shown in equation (20):

[0099]

[0100] In the equation, E[·] represents the calculation of expectation; R n,b represents the covariance matrix corresponding to the snapshot data vector b n,k The superscript H represents the conjugate transpose.

[0101] In exemplary embodiments, the result of the eigenvalue decomposition corresponding to the receiving array includes a plurality of eigenvalues and an eigenvector corresponding to each eigenvalue; and the noise subspace corresponding to each receiving array is determined based on the result of the eigenvalue decomposition corresponding to each receiving array, respectively, including:

[0102] The eigenvalues corresponding to the receiving array are sorted, and a target vector is determined from the eigenvectors corresponding to the receiving array according to the sorting result.​

[0103] Based on the target vector, the noise subspace corresponding to the receiving array is determined.

[0104] In this embodiment, for any receiving array in the MIMO radar system, the eigenvalue decomposition of the covariance matrix corresponding to that receiving array can be performed using existing matrix eigenvalue decomposition methods. The result of eigenvalue decomposition can include NL eigenvalues ​​and NL eigenvectors, with a one-to-one correspondence between the NL eigenvalues ​​and NL eigenvectors. In determining the noise subspace corresponding to the receiving array, the eigenvalues ​​corresponding to the receiving array can be sorted in descending order, and the target vector can be determined from the NL eigenvectors corresponding to the receiving array based on the sorting result.

[0105] Among these, based on the ranking results, the top PM... T Construct the signal subspace using the eigenvectors corresponding to the eigenvalues, and then use the remaining NL-PM... T Each feature vector is used as a target vector to determine the noise subspace corresponding to the receiving array. The signal subspace and noise subspace can be specifically shown in equations (21) and (22):

[0106]

[0107]

[0108] In the formula, and These represent the signal subspace and noise subspace corresponding to the nth receiving array, respectively; λ i (1≤i≤NL) is the i-th eigenvector in the eigenvalue decomposition result corresponding to the n-th receiving array.

[0109] In practice, for any receiving array in the MIMO radar system, during the eigenvalue decomposition of the covariance matrix corresponding to that receiving array, the updated baseband signal vector r in equation (19) can be used as a basis. n,k The covariance matrix corresponding to the receiving array is obtained by estimation methods, such as maximum likelihood estimation. The covariance matrix corresponding to the nth receiving array is then calculated. Maximum likelihood estimation results It can be shown in equation (23):

[0110]

[0111] In an exemplary embodiment, determining the estimation results of the motion parameters of each target under test based on the noise subspace corresponding to each of the receiving arrays and the array response matrix includes:

[0112] Based on the orthogonality between each column vector in the array response matrix corresponding to each receiving array and the corresponding noise subspace, the estimation results of the motion parameters of each target under test are determined.

[0113] In this embodiment, for the nth receiving array in each receiving array of the MIMO radar system, its corresponding array response matrix Q n For NL×M T A P-dimensional matrix, the j-th (1≤j≤M) T P) column vectors q n,j It can be shown in equation (24):

[0114] q n,j =Θ n,j' F n,j' s n,j”,j' (twenty four)

[0115] In the formula, j” = mod(j-1, P) + 1, This represents the rounding up operation, and mod(j-1,P) represents the remainder operation of (j-1) / P. According to equation (24), when j=(p-1)MT+1,…,pMT, the corresponding M is... T In each column vector, the value of each column vector depends only on the motion parameters of the p-th target and is not affected by the motion parameters of other targets. That is, for Q... n Any column vector in Q can be a function that depends only on the motion parameters of one of the multiple targets to be measured. Therefore, Q... n It can be expressed as equation (25):

[0116]

[0117] That is, Q n It can include P subarrays, each subarray including M T column vectors, for any subarray Q n,p This subarray is only related to the motion parameters of the p-th target, as shown in equation (26):

[0118]

[0119] In the formula, q n,m,p Let be the column vector in the subarray corresponding to the p-th target being tested, which is associated with the n-th receiving array and the m-th transmitting antenna.

[0120] In practice, the estimation results of the motion parameters of each target can be determined based on the orthogonality between the column vectors in the array response matrix corresponding to each receiving array and the noise subspace corresponding to the corresponding receiving array. For example, the array response matrix Q corresponding to the nth receiving array... n The column vector q in n,j Noise subspace corresponding to the nth receiving array Orthogonal, as shown in equation (27):

[0121]

[0122] In an exemplary embodiment, determining the estimation results of the motion parameters of each target under test based on the orthogonality between each column vector in the array response matrix corresponding to each receiving array and the corresponding noise subspace of the receiving array includes:

[0123] A cost function is constructed based on each column vector in the array response matrix corresponding to each of the receiving arrays and the noise subspace corresponding to each of the receiving arrays; wherein the cost function includes a set of motion parameter variables;

[0124] Based on the output value of the cost function within a preset range of motion parameters of each of the targets under test, the estimation result of the motion parameters of each target under test is determined.

[0125] In this embodiment, the actual noise level is taken into account. The calculation results are affected, therefore, q n,j and Since they are not perfectly orthogonal, a cost function can be constructed based on each column vector in the array response matrix corresponding to each receiving array and the noise subspace corresponding to each receiving array. The estimation results of the motion parameters of each target under test can be determined based on the cost function to ensure the accuracy of the estimation results of the motion parameters of each target under test.

[0126] The cost function is a function of a set of motion parameters, which may include the initial position and velocity. Therefore, the variables of the cost function include an initial position variable p. u and a velocity variable v u In a two-dimensional multi-object scenario, p u =[x u ,y u ], x u and y u The initial position variables p are respectively u In the x-axis and y-axis components, v u =[v xu ,v yu ], vxu and v yu are velocity variables v u are components on x-axis and y-axis. Thus, in a two-dimensional multi-target scenario, the cost function is a function affected by four variables (i.e., x u , y u , v xu and v yu ).

[0127] As an optional implementation, constructing the cost function based on each column vector in the array response matrix corresponding to each receiving array and the noise subspace corresponding to each receiving array comprises:

[0128] constructing a column vector function model based on each column vector in the array response matrix corresponding to each receiving array and the motion parameter variable; wherein the column vector function model is a function model irrelevant to each target to be detected;

[0129] constructing the cost function based on the column vector function model and the noise subspace corresponding to each receiving array.

[0130] In the implementation, since each subarray of Q n is only related to the motion parameter of the corresponding target to be detected, thus, the motion parameter of each column vector in any subarray Q n of Q n,p can be replaced by a motion parameter variable, thereby obtaining M T column vector function models corresponding to the nth receiving array. For example, for any vector function model q of the nth receiving array, the motion parameters x n,m,p , y p , v p and v xp in q yp can be replaced by x u , y u , v xu and v yu respectively to obtain That is, the column vector function model is only related to the receiving array and the transmitting antenna, and is irrelevant to each target to be detected.

[0131] The cost function L(p u , v u ) can be shown as formula (28):

[0132]

[0133] wherein, since the cost function L(p u , v u) is a function affected by four variables (i.e., x u , y u , v xu and v yu ), when the values of the four variables x u , y u , v xu and v yu are exactly the same as the motion parameters of a to-be-detected target, the column vector constructed by the values of the four variables is orthogonal to the noise subspace , and thus the value of the cost function L(p u , v u ) will appear a local peak. Therefore, the motion parameters of multiple to-be-detected targets can be effectively decoupled in the estimation process, and the search dimension of the embodiment method is 4, and the calculation load will not increase with the number of to-be-detected targets.

[0134] For any to-be-detected target in the multiple to-be-detected targets, in the process of determining the estimation result of the motion parameters of the to-be-detected target, a four-dimensional parameter space can be constructed based on the preset value range of the motion parameters of the to-be-detected target, and the four-dimensional parameter space is grid divided to obtain parameter values represented by each grid (for example, for any grid in the grid division result, the parameter value corresponding to the center point of the grid can be taken as the parameter value represented by the grid); the parameter values represented by each grid are respectively input into the cost function, and the parameter value corresponding to the maximum output value of the cost function is taken as the estimation result of the motion parameters of the to-be-detected target, thereby effectively ensuring the accuracy of the estimation result of the motion parameters of each to-be-detected target.

[0135] In summary, by using the cost function shown in formula (28) in the two-dimensional multi-target scene through the embodiment method, the direct estimation of the motion parameters of multiple to-be-detected targets can be realized by searching for the peak value in the four-dimensional (i.e., x u , y u , v xu and v yu ) parameter space, thereby avoiding the problem of excessive calculation load in the multi-target scene.

[0136] The specific effects of the multi-target motion parameter estimation method based on the MIMO radar of the present application are described in detail below by way of example.

[0137] It is assumed that the MIMO radar system contains 4 transmitting antennas and 4 receiving arrays, and each receiving array contains 3 receiving antennas with a half-wavelength spacing. For the sake of simplicity, it is considered that the first receiving antenna in the receiving array can realize the functions of receiving and transmitting at the same time (i.e., the case of receiving and transmitting in one body). Figure 2A simulation scenario illustrating motion parameter estimation for multiple targets based on a MIMO radar system is provided. Four receiving arrays (transmitter nodes) are located at (4000, -4000), (4000, 4000), (-4000, 4000), and (-4000, -4000), respectively. Two targets, Target 1 and Target 2, are included. Their initial positions are (0, 1000) and (500, 0), respectively, and their velocities v1 and v2 are (40, 20) and (-50, -30), respectively. The units for position and velocity are meters and meters per second, respectively. For convenience, the direct parameter estimation method based on the maximum likelihood criterion is abbreviated as ML-DPD, while the direct parameter estimation method of this invention is abbreviated as SE-DPD.

[0138] The operating parameters of the MIMO radar system are shown in Table 1. The transmitted signal used is a continuous phase-coded signal generated based on a sequential quadratic programming algorithm, where the sum of the sidelobe peaks of the autocorrelation function and the peak values ​​of the cross-correlation function is minimized during the optimization process. According to the values ​​of code length, sampling frequency, and pulse width in Table 1, the number of sampling points per symbol is 2 when using the phase-coded signal.

[0139] Table 1

[0140]

[0141]

[0142] Figure 3 The example demonstrates a comparison of the estimation accuracy of the initial position of target 1 between the method of this invention and the ML-DPD estimation method in a multi-target scenario. Figure 4 The example demonstrates a comparison of the accuracy of the proposed method in estimating the velocity of target 1 with that of the ML-DPD estimation method in a multi-target scenario. Figure 5 The example demonstrates a comparison of the estimation accuracy of the initial position of target 2 between the method of this invention and the ML-DPD estimation method in a multi-target scenario. Figure 6 The example demonstrates a comparison of the accuracy of the proposed method in estimating the velocity of target 2 with that of the ML-DPD estimation method in a multi-target scenario. Figures 3-6 In the figure, the horizontal axis represents SNR (Signal-to-Noise Ratio), and the vertical axis represents rmse (Root Mean Square Error).

[0143] Depend on Figures 3-6 It can be seen that the estimation performance of the ML-DPD estimation method is slightly better than that of the method of this invention in multi-objective scenarios. However, the search dimension of the ML-DPD estimation method in multi-objective scenarios is 4. P, P is the number of the to-be-tested targets, and the search dimension of the method in the multi-target scene is 4 and does not change with the number of the to-be-tested targets, so that the slight improvement of the estimation accuracy of the ML-DPD estimation method is obtained on the premise of exponential growth of the calculation amount, and therefore, the method can effectively reduce the calculation load under the premise of guaranteeing the estimation accuracy in the multi-target scene.

[0144] The multi-target motion parameter estimation device based on the MIMO radar provided by the application is described below, and the multi-target motion parameter estimation device based on the MIMO radar described below can be correspondingly referred to the multi-target motion parameter estimation method based on the MIMO radar described above. As shown in Figure 7 The multi-target motion parameter estimation device based on the MIMO radar provided by the application is described below, and the multi-target motion parameter estimation device based on the MIMO radar described below can be correspondingly referred to the multi-target motion parameter estimation method based on the MIMO radar described above. As shown in

[0145] The first processing module 701 is configured to perform down-conversion processing on the echo signals observed by each receiving array of the MIMO radar system to obtain an identification result of the baseband signal corresponding to each receiving array; wherein the echo signal observed by each receiving array is formed by superimposing the signals reflected to the receiving array by multiple to-be-tested targets;

[0146] The second processing module 702 is configured to construct a covariance matrix corresponding to each receiving array based on the identification result of the baseband signal corresponding to each receiving array;

[0147] The third processing module 703 is configured to perform eigenvalue decomposition on the covariance matrix corresponding to each receiving array, and determine a noise subspace corresponding to each receiving array based on the result of the eigenvalue decomposition corresponding to each receiving array;

[0148] The fourth processing module 704 is configured to determine an estimation result of the motion parameter of each to-be-tested target based on the noise subspace corresponding to each receiving array and an array response matrix; wherein the array response matrix comprises the array response characteristics of the receiving array and the signal characteristics of the echo signal observed by the receiving array.

[0149] In the exemplary embodiment, the second processing module 702 is specifically configured to:

[0150] perform signal discretization processing on the identification result of the baseband signal corresponding to the receiving array to obtain a baseband signal vector corresponding to the receiving array; wherein the baseband signal vector comprises an expected signal matrix and a noise signal matrix;

[0151] perform subspace expansion on the expected signal matrix to obtain an expected signal expansion matrix;

[0152] updating the baseband signal vector based on the desired signal expansion matrix, and constructing a covariance matrix corresponding to the receiving array based on the updated baseband signal vector.

[0153] In an example embodiment, the desired signal matrix comprises a steering vector term, a Doppler frequency offset term, a baseband signal term, a time delay term, and a scattering coefficient term.

[0154] The second processing module 702 is specifically configured to:

[0155] determine an array response matrix corresponding to the desired signal matrix based on the steering vector term, the Doppler frequency offset term, and the baseband signal term, and determine a snapshot data vector corresponding to the desired signal matrix based on the time delay term and the scattering coefficient term.

[0156] determine the desired signal expansion matrix based on the array response matrix and the snapshot data vector; wherein the steering vector term, the Doppler frequency offset term, and the time delay term are all constructed based on the motion parameter term of each of the to-be-tested targets.

[0157] In an example embodiment, the result of the eigenvalue decomposition corresponding to the receiving array comprises a plurality of eigenvalues and an eigenvector corresponding to each of the eigenvalues; the third processing module 703 is specifically configured to:

[0158] sort each of the eigenvalues corresponding to the receiving array, and determine a target vector from each of the eigenvectors corresponding to the receiving array according to the sorting result.

[0159] determine the noise subspace corresponding to the receiving array based on the target vector.

[0160] In an example embodiment, the fourth processing module 704 is specifically configured to:

[0161] determine an estimation result of the motion parameter of each of the to-be-tested targets based on the orthogonal characteristic of each column vector in the array response matrix corresponding to each of the receiving arrays and the noise subspace corresponding to the corresponding receiving array.

[0162] In an example embodiment, the fourth processing module 704 is specifically configured to:

[0163] construct a cost function based on each of the column vectors in the array response matrix corresponding to each of the receiving arrays and the noise subspace corresponding to each of the receiving arrays; wherein the cost function comprises a group of motion parameter variables.

[0164] Determine an estimation result of the motion parameter of each of the to-be-tested targets based on output values of the cost function in a preset value interval of the motion parameter of each of the to-be-tested targets.

[0165] In an example embodiment, the fourth processing module 704 is specifically configured to:

[0166] Construct a column vector function model based on each of the column vectors in the array response matrix corresponding to each of the receiving arrays and the motion parameter variable, wherein the column vector function model is a function model irrelevant to each of the to-be-tested targets;

[0167] Construct the cost function based on the column vector function model and the noise subspace corresponding to each of the receiving arrays.

[0168] Figure 8 An example of a schematic diagram of the physical structure of an electronic device is shown in Figure 8 The electronic device can include a processor 801, a communications interface 802, a memory 803, and a communications bus 804, wherein the processor 801, the communications interface 802, and the memory 803 complete mutual communication through the communications bus 804. The processor 801 can invoke a logical instruction in the memory 803 to execute a multi-target motion parameter estimation method based on a MIMO radar, which includes: performing down-conversion processing on echo signals observed by each receiving array of a MIMO radar system to obtain an identification result of a baseband signal corresponding to each receiving array; wherein the echo signal observed by each receiving array is formed by superimposing signals reflected to the receiving array by multiple to-be-tested targets;

[0169] Construct a covariance matrix corresponding to each receiving array based on the identification result of the baseband signal corresponding to each receiving array;

[0170] Perform eigenvalue decomposition on the covariance matrix corresponding to each receiving array, and determine a noise subspace corresponding to each receiving array based on a result of the eigenvalue decomposition corresponding to each receiving array;

[0171] Determine an estimation result of the motion parameter of each of the to-be-tested targets based on the noise subspace corresponding to each receiving array and an array response matrix, wherein the array response matrix includes array response characteristics of the receiving array and signal characteristics of the echo signal observed by the receiving array.

[0172] Moreover, the logic instructions in the memory 803 described above can be implemented in the form of software functional units and sold or used as independent products, and can be stored in a computer readable storage medium. Based on such understanding, the technical solutions of the present application essentially or the parts that make contributions to the prior art or parts of the technical solutions can be embodied in the form of a software product. The computer software product is stored in a storage medium, and includes several instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present application. The aforementioned storage medium includes: a U disk, a mobile hard disk, a read-only memory (ROM, Read-Only Memory), a random access memory (RAM, Random Access Memory), a magnetic disk or an optical disk, and various media that can store program codes.

[0173] In another aspect, the present application also provides a computer program product, which comprises a computer program, the computer program can be stored on a non-transitory computer readable storage medium, and the computer program can be executed by a processor to enable a computer to execute the MIMO radar based multi-target motion parameter estimation method provided by the above-mentioned methods. The method comprises: performing down-conversion processing on the echo signals observed by each receiving array of a MIMO radar system to obtain the identification results of the baseband signals corresponding to each receiving array; wherein the echo signals observed by each receiving array are formed by superimposing the signals reflected by multiple to-be-measured targets to the receiving array.

[0174] Based on the identification results of the baseband signals corresponding to each receiving array, a covariance matrix corresponding to each receiving array is constructed respectively;

[0175] Eigenvalue decomposition is performed on the covariance matrices corresponding to each receiving array respectively, and based on the results of the eigenvalue decomposition corresponding to each receiving array, a noise subspace corresponding to each receiving array is determined respectively;

[0176] Based on the noise subspaces corresponding to each receiving array and an array response matrix, the estimation results of the motion parameters of each to-be-measured target are determined; wherein the array response matrix comprises the array response characteristics of the receiving array and the signal characteristics of the echo signals observed by the receiving array.

[0177] In yet another aspect, the present application also provides a non-transitory computer readable storage medium having stored thereon a computer program, which, when executed by a processor, implements a method for estimating motion parameters of multiple targets based on a MIMO radar, the method comprising: performing down-conversion processing on echo signals observed by each receiving array of a MIMO radar system to obtain an identification result of a baseband signal corresponding to each receiving array; wherein the echo signals observed by each receiving array are formed by superimposing signals reflected by multiple targets to be measured to the receiving array;

[0178] Based on the identification result of the baseband signal corresponding to each receiving array, a covariance matrix corresponding to each receiving array is constructed respectively;

[0179] Eigenvalue decomposition is performed on the covariance matrix corresponding to each receiving array respectively, and based on the result of the eigenvalue decomposition corresponding to each receiving array, a noise subspace corresponding to each receiving array is determined respectively;

[0180] Based on the noise subspace corresponding to each receiving array and an array response matrix, an estimation result of the motion parameters of each target to be measured is determined; wherein the array response matrix comprises array response characteristics of the receiving array and signal characteristics of the echo signals observed by the receiving array.

[0181] The device embodiments described above are merely illustrative, wherein the units illustrated as separate components can or can not be physically separated, and the components illustrated as units can or can not be physical units, i.e., they can be located in one place, or distributed on multiple network units. Part or all of the modules can be selected to achieve the purpose of the embodiment scheme according to actual needs. Those skilled in the art can understand and implement it without creative labor.

[0182] From the above description of the embodiments, those skilled in the art can clearly understand that the embodiments can be realized by means of software plus necessary universal hardware platforms, and of course, can also be realized by hardware. Based on such understanding, the above technical solutions, essentially or in other words, the part that contributes to the prior art, can be embodied in the form of a software product, which can be stored in a computer readable storage medium, such as a ROM / RAM, a magnetic disk, an optical disk, etc., and includes a number of instructions to make a computer device (which can be a personal computer, a server, or a network device, etc.) execute the methods described in each embodiment or some parts of the embodiments.

[0183] It should be pointed out finally that the above embodiments are only used to illustrate the technical solutions of the present application, but not to limit the same; and although the present application has been described in detail with reference to the foregoing embodiments, it should be appreciated by those skilled in the art that the technical solutions recorded in the foregoing embodiments can be modified, or some technical features thereof can be replaced equivalently; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A method for multi-target motion parameter estimation based on MIMO radar, characterized in that, The method comprises: down-converting echo signals observed by each receiving array of a MIMO radar system to obtain identification results of baseband signals corresponding to each receiving array; wherein the echo signals observed by each receiving array are formed by superimposing signals reflected by multiple to-be-measured targets to the receiving array; based on the identification results of the baseband signals corresponding to each receiving array, constructing a covariance matrix corresponding to each receiving array respectively; performing eigenvalue decomposition on the covariance matrix corresponding to each receiving array respectively, and determining a noise subspace corresponding to each receiving array based on the results of the eigenvalue decomposition corresponding to each receiving array; based on the noise subspace corresponding to each receiving array and an array response matrix, determining an estimation result of the motion parameters of each to-be-measured target; wherein the array response matrix comprises array response characteristics of the receiving array and signal characteristics of the echo signals observed by the receiving array; wherein, based on the identification results of the baseband signals corresponding to each receiving array, constructing a covariance matrix corresponding to each receiving array respectively, comprises: performing signal discretization processing on the identification results of the baseband signals corresponding to the receiving array to obtain a baseband signal vector corresponding to the receiving array; wherein the baseband signal vector comprises an expected signal matrix and a noise signal matrix; performing subspace expansion on the expected signal matrix to obtain an expected signal expansion matrix; updating the baseband signal vector based on the expected signal expansion matrix, and constructing the covariance matrix corresponding to the receiving array based on the updated baseband signal vector; wherein, the expected signal matrix comprises a steering vector term, a Doppler frequency offset term, a baseband signal term, a time delay term and a scattering coefficient term; the subspace expansion on the expected signal matrix to obtain an expected signal expansion matrix, comprises: based on the steering vector term, the Doppler frequency offset term and the baseband signal term, determining an array response matrix corresponding to the expected signal matrix, and based on the time delay term and the scattering coefficient term, determining a snapshot data vector corresponding to the expected signal matrix; determining the expected signal expansion matrix based on the array response matrix and the snapshot data vector; wherein the steering vector term, the Doppler frequency offset term and the time delay term are constructed based on the motion parameter term of each to-be-measured target.

2. The method of claim 1, wherein, the results of the eigenvalue decomposition corresponding to the receiving array comprise a plurality of eigenvalues and eigenvectors corresponding to each eigenvalue; based on the results of the eigenvalue decomposition corresponding to each receiving array, determining a noise subspace corresponding to each receiving array, comprises: sorting each eigenvalue corresponding to the receiving array, and determining a target vector from the eigenvectors corresponding to the receiving array according to the sorting result; based on the target vector, determining the noise subspace corresponding to the receiving array.

3. The method of claim 1, wherein, based on the noise subspace corresponding to each receiving array and the array response matrix, determining an estimation result of the motion parameters of each to-be-measured target, comprises: The estimation result of the motion parameter of each of the to-be-detected targets is determined based on orthogonality of each column vector in the array response matrix corresponding to each of the receiving arrays and the noise subspace corresponding to the receiving array.

4. The method of claim 3, wherein, The estimation result of the motion parameter of each of the to-be-detected targets is determined based on orthogonality of each column vector in the array response matrix corresponding to each of the receiving arrays and the noise subspace corresponding to the receiving array. The cost function is constructed based on each column vector in the array response matrix corresponding to each of the receiving arrays and the noise subspace corresponding to each of the receiving arrays; wherein the cost function includes a set of motion parameter variables. The estimation result of the motion parameter of each of the to-be-detected targets is determined based on output values of the cost function in a preset value interval of the motion parameter of each of the to-be-detected targets.

5. The method of claim 3, wherein, The cost function is constructed based on each column vector in the array response matrix corresponding to each of the receiving arrays and the noise subspace corresponding to each of the receiving arrays, including: A column vector function model is constructed based on each column vector in the array response matrix corresponding to each of the receiving arrays and the motion parameter variable; wherein the column vector function model is a function model irrelevant to each of the to-be-detected targets; The cost function is constructed based on the column vector function model and the noise subspace corresponding to each of the receiving arrays.

6. A multi-target motion parameter estimation apparatus based on MIMO radar, characterized by, including: The first processing module is configured to perform down-conversion processing on echo signals observed by each receiving array of a MIMO radar system to obtain an identification result of a baseband signal corresponding to each receiving array; wherein the echo signal observed by each receiving array is formed by superimposing signals reflected to the receiving array by multiple to-be-detected targets. The second processing module is configured to construct a covariance matrix corresponding to each receiving array based on the identification result of the baseband signal corresponding to each receiving array. The third processing module is configured to perform eigenvalue decomposition on the covariance matrix corresponding to each receiving array, respectively, and determine a noise subspace corresponding to each receiving array based on a result of the eigenvalue decomposition corresponding to each receiving array. The fourth processing module is configured to determine an estimation result of a motion parameter of each to-be-detected target based on the noise subspace corresponding to each receiving array and an array response matrix; wherein the array response matrix includes array response characteristics of the receiving array and signal characteristics of the echo signal observed by the receiving array. The second processing module is configured to perform signal discretization processing on the identification result of the baseband signal corresponding to the receiving array to obtain a baseband signal vector corresponding to the receiving array; wherein the baseband signal vector includes an expected signal matrix and a noise signal matrix; perform subspace expansion on the expected signal matrix to obtain an expected signal expansion matrix; update the baseband signal vector based on the expected signal expansion matrix, and construct the covariance matrix corresponding to the receiving array based on the updated baseband signal vector. The expected signal matrix comprises a steering vector item, a Doppler frequency offset item, a baseband signal item, a time delay item and a scattering coefficient item. The second processing module is configured to determine an array response matrix corresponding to the expected signal matrix based on the steering vector item, the Doppler frequency offset item and the baseband signal item, and determine a snapshot data vector corresponding to the expected signal matrix based on the time delay item and the scattering coefficient item; determine the expected signal extension matrix based on the array response matrix and the snapshot data vector; wherein the steering vector item, the Doppler frequency offset item and the time delay item are all constructed based on a motion parameter item of each of the to-be-measured targets.

7. An electronic device comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, characterized in that, The processor executes the program to implement the MIMO radar-based multi-target motion parameter estimation method according to any one of claims 1 to 5.

8. A non-transitory computer-readable storage medium having stored thereon a computer program, characterized in that, The computer program is executed by the processor to implement the MIMO radar-based multi-target motion parameter estimation method according to any one of claims 1 to 5.

Images