Multi-target parameter dimension reduction estimation method and device for frequency control array radar

Abstract

The application relates to the technical field of radars, and provides a multi-target parameter dimension reduction estimation method and device for a frequency control array radar, which comprises the following steps: determining a receiving signal matrix of the frequency control array radar based on a transmitting steering vector, a receiving steering vector of each target and a transmitting signal of the frequency control array radar at each snapshot moment; solving the receiving signal matrix by using a high-order singular value decomposition method to obtain a noise subspace matrix of the receiving signal matrix; constructing a multi-target parameter dimension reduction estimation model based on the noise subspace matrix and the transmitting steering vector and the receiving steering vector of each target; solving the multi-target parameter dimension reduction estimation model by using a polynomial root-finding method to estimate an angle estimation value of each target; and estimating a distance estimation value of each target corresponding to the angle estimation value of each target based on the angle estimation value of each target. The application not only reduces the calculation complexity under the condition of few snapshots, but also improves the precision of parameter estimation.

Description

Technical Field

[0001] The present invention relates to the field of radar technology, and in particular to a multi-target parameter dimensionality reduction estimation method and device for frequency-controlled array radar. Background Art

[0002] In practical radar applications, simultaneous detection of multiple targets (such as multiple aircraft) is often necessary. However, when the target angles (i.e., the horizontal azimuth angles of the targets relative to the radar) are close or identical, traditional radar systems such as phased array radars cannot effectively distinguish between multiple targets due to their fixed beam patterns. Compared to phased array radars, frequency-controlled array radars operate by using different carriers between each array element, with a fixed frequency offset between each carrier. The receiver can extract range and angle information from the target signal, forming a two-dimensional beam pattern with range-angle decoupling. This effectively distinguishes multiple targets and performs two-dimensional parameter estimation of the target's range and angle.

[0003] However, frequency-steering array radars suffer from signal non-stationarity due to frequency offsets between array elements, which reduces the number of available snapshots—that is, the number of independent signal samples. Furthermore, dynamic target scenarios and strong interference also reduce the number of valid snapshots. When the number of available snapshots is extremely small (less than 10) and the target angles are similar or identical, existing dimensionality reduction estimation algorithms for frequency-steering array radars are prone to missing targets, resulting in low accuracy in multi-target parameter estimation. Summary of the Invention

[0004] The present invention provides a multi-target parameter dimensionality reduction estimation method and device for a frequency-controlled array radar, which is used to solve the problem in the prior art that when the number of available snapshots of the frequency-controlled array radar is very small and the angles of the targets are close or the same, the dimensionality reduction estimation algorithm used for the frequency-controlled array radar is prone to lose the target, resulting in low estimation accuracy of the multi-target parameter estimation.

[0005] The present invention provides a multi-target parameter dimensionality reduction estimation method for a frequency-controlled array radar, comprising:

[0006] Determine the receiving signal matrix of the frequency-steering array radar based on the transmitting steering vector and receiving steering vector of the frequency-steering array radar for each target and the transmitting signal of the frequency-steering array radar at each snapshot moment;

[0007] Solving the received signal matrix using a high-order singular value decomposition method to obtain a noise subspace matrix of the received signal matrix;

[0008] Based on the noise subspace matrix and the transmission steering vector and the reception steering vector of the frequency-controlled array radar for each target, a multi-target parameter dimensionality reduction estimation model is constructed;

[0009] A polynomial root-finding method is used to solve the multi-objective parameter dimensionality reduction estimation model to estimate the angle estimation value of each target;

[0010] Based on the angle estimation value of each target, the multi-target parameter dimensionality reduction estimation model is solved by using a polynomial rooting method to estimate the distance estimation value of each target corresponding to the angle estimation value of each target.

[0011] According to the present invention, a multi-target parameter dimensionality reduction estimation method for a frequency-steering array radar is provided. The method determines a received signal matrix of the frequency-steering array radar based on the transmission steering vector and reception steering vector of each target by the frequency-steering array radar and the transmission signal of the frequency-steering array radar at each snapshot time. The method includes:

[0012] Based on the transmit and receive steering vectors of the frequency-controlled array radar to each target, the joint transmit and receive array manifold matrix is ​​determined;

[0013] Based on the transmit signals at each snapshot moment, a preset noise vector, and the joint transmit-receive array manifold matrix, a receive signal matrix of the frequency-controlled array radar is determined, wherein the transmit signals at each snapshot moment all satisfy an orthogonality condition.

[0014] According to a multi-target parameter dimensionality reduction estimation method for a frequency-controlled array radar provided by the present invention, a high-order singular value decomposition method is used to solve the received signal matrix to obtain a noise subspace matrix of the received signal matrix, including:

[0015] Reconstructing the received signal matrix into a third-order tensor;

[0016] performing high-order singular value decomposition on the third-order tensor and performing truncation processing on the decomposed third-order tensor to obtain a signal subspace matrix based on the third-order tensor;

[0017] The signal subspace matrix is ​​sequentially subjected to normalization and orthogonal projection to obtain the noise subspace matrix.

[0018] According to the present invention, a multi-target parameter dimensionality reduction estimation method for a frequency-steering array radar is provided. Based on the noise subspace matrix and the transmission steering vector and reception steering vector of each target by the frequency-steering array radar, a multi-target parameter dimensionality reduction estimation model is constructed, comprising:

[0019] Determine the total steering vector based on the transmit steering vector and receive steering vector of each target by the frequency-controlled array radar;

[0020] constructing a cost function based on the total steering vector and the noise subspace matrix, the cost function including a distance portion of the transmit steering vector and a sign matrix, the sign matrix being determined by an angle portion of the transmit steering vector and a receive steering vector;

[0021] construct a Lagrange function based on the cost function, wherein a Lagrange multiplier in the Lagrange function is determined based on a partial derivative of the Lagrange function with respect to a distance part in the transmit steering vector being equal to 0;

[0022] determine an estimated value of the distance part in the transmit steering vector based on the Lagrange function;

[0023] replace the distance part in the transmit steering vector in the cost function with the estimated value of the distance part in the transmit steering vector to obtain a replaced cost function;

[0024] determine that the replaced cost function that is minimized is the multi-target parameter reduced-dimension estimation model.

[0025] According to the multi-target parameter reduced-dimension estimation method for the frequency-controlled array radar provided by the present application, the polynomial root-finding method is used to solve the multi-target parameter reduced-dimension estimation model to estimate the angle estimation values of the targets, which comprises:

[0026] when the multi-target parameter reduced-dimension estimation model reaches the minimum, the determinant of the sign matrix is set to 0;

[0027] the fast Fourier transform is used to calculate the determinant of the sign matrix, and when the determinant of the sign matrix is 0, a 2M(N-1) order polynomial is obtained, wherein M represents the number of transmit elements of the frequency-controlled array radar, and N represents the number of receive elements of the frequency-controlled array radar;

[0028] the 2M(N-1) order polynomial is solved to obtain a plurality of first complex roots;

[0029] K first target complex roots closest to the unit circle are selected from the plurality of first complex roots, and K represents the number of targets;

[0030] based on the K first target complex roots and the spacing between the receive elements of the frequency-controlled array radar, the angle estimation values of the targets are estimated.

[0031] According to the multi-target parameter reduced-dimension estimation method for the frequency-controlled array radar provided by the present application, based on the angle estimation values of the targets, the polynomial root-finding method is used to solve the multi-target parameter reduced-dimension estimation model to estimate the distance estimation values of the targets corresponding to the angle estimation values of the targets, which comprises:

[0032] the angle estimation values of the targets are respectively substituted into the multi-target parameter reduced-dimension estimation model;

[0033] for the angle estimation value of any target, a second complex root of the polynomial when the corresponding multi-target parameter reduced-dimension estimation model is 0 is solved;

[0034] An estimated distance value of each target is estimated based on the second complex-valued root and the frequency offset of the frequency-controlled array radar.

[0035] The present invention also provides a multi-target parameter dimensionality reduction estimation device for a frequency-controlled array radar, comprising:

[0036] a receiving signal matrix determination module, configured to determine the receiving signal matrix of the frequency-steering array radar based on the transmitting steering vector and receiving steering vector of the frequency-steering array radar for each target and the transmitting signal of the frequency-steering array radar at each snapshot;

[0037] A receiving signal matrix solving module is used to solve the receiving signal matrix using a high-order singular value decomposition method to obtain a noise subspace matrix of the receiving signal matrix;

[0038] A model building module is used to build a multi-target parameter dimensionality reduction estimation model based on the noise subspace matrix and the transmission steering vector and reception steering vector of the frequency-controlled array radar for each target;

[0039] An angle estimation module, used to solve the multi-objective parameter dimensionality reduction estimation model using a polynomial root-finding method to estimate the angle estimation value of each target;

[0040] The distance estimation module is used to solve the multi-target parameter dimensionality reduction estimation model based on the angle estimation value of each target by using the polynomial root-finding method to estimate the distance estimation value of each target corresponding to the angle estimation value of each target.

[0041] The present invention also provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the program, it implements any of the above-described multi-target parameter dimensionality reduction estimation methods for frequency-controlled array radars.

[0042] The present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, the method for multi-target parameter dimensionality reduction estimation for a frequency-controlled array radar as described above is implemented.

[0043] The present invention also provides a computer program product, comprising a computer program, which, when executed by a processor, implements any of the above-described multi-target parameter dimensionality reduction estimation methods for frequency-controlled array radar.

[0044] The present invention provides a multi-target parameter dimensionality reduction estimation method and device for a frequency-controlled array radar. The method determines a received signal matrix of the frequency-controlled array radar based on the transmission steering vector and reception steering vector of each target by the frequency-controlled array radar and the transmission signal of the frequency-controlled array radar at each snapshot moment; solves the received signal matrix using a high-order singular value decomposition method to obtain a noise subspace matrix of the received signal matrix; constructs a multi-target parameter dimensionality reduction estimation model based on the noise subspace matrix and the transmission steering vector and reception steering vector of each target by the frequency-controlled array radar; solves the multi-target parameter dimensionality reduction estimation model using a polynomial rooting method to estimate an angle estimate of each target; and solves the multi-target parameter dimensionality reduction estimation model based on the angle estimate of each target using a polynomial rooting method to estimate a distance estimate of each target corresponding to the angle estimate of each target. That is, the high-order singular value decomposition method is used to obtain the noise subspace of the frequency-controlled array radar receiving signal matrix. Using the noise subspace, the polynomial root-finding method is used to solve the dimensionality reduction multiple signal classification method, which can effectively estimate the distance and angle of multiple targets. Under the condition of few snapshots, it not only reduces the computational complexity but also improves the accuracy of parameter estimation. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] In order to more clearly illustrate the technical solutions in the present invention or the prior art, a brief introduction is given below to the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0046] Figure 1 It is a flow chart of a multi-target parameter dimensionality reduction estimation method for a frequency-controlled array radar provided by the present invention.

[0047] Figure 2 The present invention provides a schematic diagram of a frequency-controlled array array structure of a frequency-controlled array radar in a multi-target parameter dimensionality reduction estimation method for a frequency-controlled array radar.

[0048] Figure 3 It is a schematic diagram showing how the computational complexity of the multi-target parameter dimensionality reduction estimation method for a frequency-controlled array radar provided by the present invention varies with the number of array elements of the frequency-controlled array radar.

[0049] Figure 4 It is a schematic diagram showing how the root mean square error between the angle estimate value and the true angle value estimated by the multi-target parameter dimensionality reduction estimation method for frequency-controlled array radar provided by the present invention changes with the signal-to-noise ratio.

[0050] Figure 5 It is a schematic diagram showing how the root mean square error between the distance estimate value and the true distance value estimated by the multi-target parameter dimensionality reduction estimation method for frequency-controlled array radar provided by the present invention changes with the signal-to-noise ratio.

[0051] Figure 6 Figure 3 is a schematic view of the root mean square error of the estimated angle value and the real angle value with the change of the signal-to-noise ratio when the multi-target angle estimation method for the frequency control array radar provided in the present application estimates the multi-targets with similar angles.

[0052] Figure 7 Figure 4 is a schematic view of the root mean square error of the estimated distance value and the real distance value with the change of the signal-to-noise ratio when the multi-target angle estimation method for the frequency control array radar provided in the present application estimates the multi-targets with similar angles.

[0053] Figure 8 Figure 5 is a schematic view of the root mean square error of the estimated angle value and the real angle value with the change of the number of fast shots when the multi-target angle estimation method for the frequency control array radar provided in the present application estimates the multi-targets with similar angles.

[0054] Figure 9 Figure 6 is a schematic view of the root mean square error of the estimated distance value and the real distance value with the change of the number of fast shots when the multi-target angle estimation method for the frequency control array radar provided in the present application estimates the multi-targets with similar angles.

[0055] Figure 10 Figure 7 is a structural schematic view of the multi-target parameter dimension reduction estimation device for the frequency control array radar provided in the present application.

[0056] Figure 11 Figure 8 is a structural schematic view of the electronic device provided in the present application. DETAILED DESCRIPTION

[0057] In order 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 with reference to the drawings in the present application. Obviously, the described embodiments are some embodiments of the present application, but not all the embodiments. Based on the embodiments in the present application, all the other embodiments obtained by those skilled in the art without creative labor fall within the protection scope of the present application.

[0058] The multi-target parameter dimension reduction estimation method for the frequency control array radar in the embodiment of the present application, as shown in FIG. 1, comprises the following steps S110 to S150. Figure 1

[0059] Step S110: determining the receiving signal matrix of the frequency control array radar based on the transmission steering vector, the receiving steering vector of each target and the transmission signal of the frequency control array radar at each fast shot moment.

[0060] Figure 2 ​The frequency-controlled array structure of the frequency-controlled array radar has M antennas at the transmitting end, i.e., M transmitting array elements, and N antennas at the receiving end, i.e., N receiving array elements. d T and d R are the transmitting array element spacing and the receiving array element spacing, respectively. θ Indicates the target angle. The carrier frequencies of each transmitting array element are , ,…, ,in, and In this step, the frequency-steering array radar's received signal matrix can be obtained by using the frequency-steering array radar's transmit steering vectors and receive steering vectors for each target, as well as the frequency-steering array radar's transmit signals at each snapshot. The receive signal matrix is ​​composed of the receive signals at each snapshot.

[0061] Step S120: Utilize high-order singular value decomposition (HOSVD) to solve the received signal matrix and obtain the noise subspace matrix of the received signal matrix. When the number of snapshots is small, traditional eigenvalue decomposition (Eigenvalue Decomposition) is insufficient to effectively separate the signal subspace from the noise subspace. Therefore, in this step, the High-Order Singular Value Decomposition (HOSVD) method is employed to solve the received signal matrix, effectively separating the signal subspace from the noise subspace and obtaining the noise subspace matrix. Furthermore, using HOSVD to process the received signal matrix can better handle noise with fewer snapshots, thereby improving parameter estimation accuracy.

[0062] Step S130: A multi-target parameter dimensionality reduction estimation model is constructed based on the noise subspace matrix and the transmit and receive steering vectors for each target from the frequency-controlled array radar. Because the transmit steering vectors contain signals related to both the angle and range components, the multi-target parameter dimensionality reduction estimation model also includes the angle and range information for each target. Subsequently, the multi-target parameter dimensionality reduction estimation model is solved using a polynomial to obtain the angle and range of each target.

[0063] Step S140: the polynomial root method is used to solve the multi-target parameter dimension reduction estimation model, so as to estimate the angle estimation value of each target. By reducing the dimension of the parameter estimation model of the frequency control array radar, the angle of the target is estimated first, and then the distance of the target is estimated at the specific angle, so that the estimation mode of two-dimensional joint distance and angle estimation is no longer used, and the calculation complexity is reduced. Moreover, the polynomial root method is used to solve the multi-target parameter dimension reduction estimation model, so that the spectral peak search is not needed, and therefore the target is not lost (the traditional dimension reduction parameter estimation method needs to perform the spectral peak search in the process of solving the model, and when the target angle or distance is close, the spectral peak may be fused to cause the loss of the target), and the parameter estimation precision is improved.

[0064] Step S150: based on the angle estimation value of each target, the polynomial root method is used to solve the multi-target parameter dimension reduction estimation model, so as to estimate the distance estimation value of each target corresponding to the angle estimation value of each target.

[0065] The multi-target parameter dimension reduction estimation method for the frequency control array radar in the embodiment uses the high-order singular value decomposition method to obtain the noise subspace of the frequency control array radar receiving signal matrix, better processes noise under the condition of few snapshots, and uses the noise subspace to solve the multi-target parameter dimension reduction estimation model by using the polynomial root method, so that the target is not lost, thereby avoiding the calculation error caused by the loss of the target under the condition of few snapshots in the traditional dimension reduction parameter estimation algorithm, improving the parameter estimation precision, and further effectively estimating the distance and angle of the multi-target.

[0066] In some embodiments, in step S110, based on the transmission steering vector, the receiving steering vector of the frequency control array radar to each target and the transmission signal of the frequency control array radar at each snapshot moment, the receiving signal matrix of the frequency control array radar is determined, including:

[0067] Step S111: based on the transmission steering vector and the receiving steering vector of the frequency control array radar to each target, a joint transmitting-receiving array manifold matrix is determined.

[0068] Suppose that there are K targets in total, the distance of the kth target is r k and the angle is θ k The transmission steering vector of each target is:

[0069] (1);

[0070] (2);

[0071] (3)。

[0072] Wherein, and They are k The distance and angle components of the launch guidance vector for each target, and[·] T represent Hadamard product operation and transpose operation respectively. ,in, d T represents the distance between transmitting elements, c represents the speed of light, λ The wavelength of the carrier center frequency, j is an imaginary unit, e is a natural constant, Represents a complex field.

[0073] No. k The receiving steering vector of a target is:

[0074] (4).

[0075] Among them, here d R Indicates the receiving array element spacing. It should be noted that: d T Can be equal to d R , d T and d R They are all designed to be half a wavelength in length, thus maximizing the performance of the frequency-controlled array radar while taking into account feasibility.

[0076] The manifold matrix A of the joint transmit-receive array is:

[0077] (5).

[0078] in, represents the Kronecker product.

[0079] Step S112: Determine a receive signal matrix of the frequency-controlled array radar based on the transmit signals at each snapshot moment, a preset noise vector, and the joint transmit-receive array manifold matrix, wherein the transmit signals at each snapshot moment satisfy an orthogonality condition.

[0080] Specifically, when the transmitted signal waveform meets the orthogonality condition (meeting the orthogonality condition can effectively separate the transmitted signal of each transmitting antenna), then after matched filtering, the first l The received signal of a snapshot can be expressed as:

[0081] (6).

[0082] in, is the vector of the transmitted signal after matched filtering, is the Gaussian noise vector, t l Indicates the l A snapshot of the corresponding moment.

[0083] If the number of snapshots is L , then the receiving signal matrix of the frequency-controlled array radar can be expressed as:

[0084] (7).

[0085] in, , .

[0086] In some embodiments, step S120, using a high-order singular value decomposition method to solve the received signal matrix to obtain a noise subspace matrix of the received signal matrix, includes:

[0087] Step S121: Reconstruct the received signal matrix into a third-order tensor . At the same time, you can get:

[0088] (8).

[0089] in, It is the modulo three expansion, which means expanding the matrix along the third dimension.

[0090] It can be understood that reconstruction is to reorganize each element of the first dimension of the MN×L dimensional received signal matrix Y into two dimensions M×N in a column-first manner, and finally obtain a three-dimensional tensor with M rows, N columns and L layers from the received signal matrix Y.

[0091] Step S122: performing high-order singular value decomposition on the third-order tensor, and performing truncation processing on the decomposed third-order tensor to obtain a signal subspace matrix based on the third-order tensor.

[0092] For a third-order tensor Performing high-order singular value decomposition yields:

[0093] (9).

[0094] in, It is a tensor The core tensor obtained by high-order singular value decomposition, 、 and They represent the modulo 1 product, modulo 2 product, and modulo 3 product respectively. Represents a third-order tensor The left singular matrix obtained by performing singular value decomposition on the matrix after modulo 1 expansion is Represents a third-order tensor The left singular matrix obtained by performing singular value decomposition on the matrix after modulo 2 expansion is Represents a third-order tensor The left singular matrix obtained by performing singular value decomposition on the matrix after modulo 3 expansion, the specific calculation formulas for U1, U2 and U3 are as follows:

[0095] (10).

[0096] Among them, n=1,2,3, Represents a third-order tensor The singular value matrix obtained by performing singular value decomposition on the matrix after the module n expansion is Represents a third-order tensor The right singular matrix obtained by performing singular value decomposition on the matrix after the module n expansion, H represents the conjugate transpose operation of the matrix.

[0097] Using truncation method to process third-order tensors , we get the signal subspace based on the third-order tensor :

[0098] (11).

[0099] in, represents the truncated core tensor, U Sn Is to intercept U n The matrix formed by the K column vectors from left to right in .

[0100] The core tensor Substituting into formula (11), we get:

[0101] (12).

[0102] The signal subspace based on the third-order tensor Perform modulo 3 expansion to obtain the signal subspace matrix U S :

[0103] (13).

[0104] in, Represents the matrix conjugation operation.

[0105] Step S123: Perform unitization and orthogonal projection on the signal subspace matrix in sequence to obtain the noise subspace matrix. Specifically, the signal subspace matrix U is transformed by the unitization operation. S , that is, the signal subspace matrix U SUnitization is performed so that the Euclidean norm of each column vector is 1, and then the noise subspace matrix U is obtained by orthogonal projection NO The orthogonal projection formula is shown in the following formula (14).

[0106] (14).

[0107] Wherein I MN denotes an MxN dimensional identity matrix, and orthogonal projection is the identity matrix minus the projection matrix of the signal subspace , so that the projection matrix of the noise subspace can be obtained. That is, the noise subspace matrix U NO can be calculated according to formula (14).

[0108] In some embodiments, step S130, based on the noise subspace matrix and the transmit steering vector and receive steering vector of each target of the frequency control array radar, a multi-target parameter dimension reduction estimation model is constructed, comprising:

[0109] Step S131: based on the transmit steering vector and receive steering vector of each target of the frequency control array radar, the total steering vector is determined.

[0110] (15).

[0111] Wherein, denotes the matrix diagonalization operation.

[0112] Step S132: based on the total steering vector and the noise subspace matrix, a cost function is constructed, the cost function includes a distance part in the transmit steering vector and a sign matrix, the sign matrix is determined by the angle part in the transmit steering vector and the receive steering vector.

[0113] Specifically, the cost function is constructed as follows:

[0114] (16).

[0115] Substitute formula (15) into formula (16) to obtain:

[0116] (17).

[0117] Wherein, denotes the sign matrix, and the sign matrix is determined by the angle part in the transmit steering vector and the receive steering vector, that is .

[0118] Minimizing the cost function is equivalent to:

[0119] (18).

[0120] wherein, , denotes the real number field.

[0121] Step S133: constructing a Lagrange function based on the cost function, wherein a Lagrange multiplier in the Lagrange function is determined based on that a partial derivative of the Lagrange function with respect to a distance part in the transmit steering vector is equal to 0.

[0122] Specifically, the Lagrange function is constructed as follows based on the cost function

[0123] (19).

[0124] μ denotes the Lagrange multiplier, and let take a partial derivative of with respect to and let the partial derivative be 0:

[0125] (20).

[0126] Based on the formula (20), the following can be obtained .

[0127] Step S134: determining an estimated value of the distance part in the transmit steering vector based on the Lagrange function. Specifically, substituting μ the formula (20) and the formula (19) obtains an estimated value of the distance part in the transmit steering vector:

[0128] (21).

[0129] Step S135: replacing the distance part in the transmit steering vector in the cost function with the estimated value of the distance part in the transmit steering vector, to obtain a replaced cost function.

[0130] Specifically, replacing the estimated value of the distance part in the transmit steering vector with in the formula (18) obtains a replaced cost function, which is expressed as follows in terms of minimizing the cost function:

[0131] (22).

[0132] Step S136: determining the minimized substitute cost function as the multi-target parameter dimension reduction estimation model. That is, formula (22) is determined as the multi-target parameter dimension reduction estimation model. In this step, for the convenience of subsequent calculation, formula (22) is rewritten to obtain formula (23), and formula (23) is the final multi-target parameter dimension reduction estimation model.

[0133] (23).

[0134] wherein, denotes the adjugate matrix, denotes the determinant of the matrix.

[0135] In this embodiment, the cost function of formula (16) utilizes the orthogonality of the total steering vector and the noise subspace. When the orthogonality is maximum, that is, the cost function is minimum, the distance and the angle substituted are the distance and the angle to be solved, which facilitates the subsequent solution of the angle and the distance of the target.

[0136] In some embodiments, step S140, a polynomial root-finding method is used to solve the multi-target parameter dimension reduction estimation model to estimate the angle estimation value of each target, including:

[0137] Step S141: when the multi-target parameter dimension reduction estimation model reaches the minimum, the determinant of the symbolic matrix is 0. Specifically, formula (23) reaches the minimum value, and the multi-target parameter dimension reduction estimation model can be equivalently represented as , is represented as a polynomial matrix about z:

[0138] (24).

[0139] wherein, , , .

[0140] Step S142: the determinant of the symbolic matrix is calculated by using the fast Fourier transform, and when the determinant of the symbolic matrix is 0, a 2M(N-1) order polynomial is obtained, wherein M represents the number of transmitting array elements of the frequency control array radar, and N represents the number of receiving array elements of the frequency control array radar. Specifically, the polynomial is a 2M(N-1) order polynomial.

[0141] Step S143: the roots of the 2M(N-1) order polynomial are obtained, that is, the roots of are obtained, and a plurality of first complex value roots are obtained.

[0142] Step S144: selecting K first target complex value roots closest to the unit circle from the plurality of first complex value roots , K represents the number of targets. The polynomial The 2M(N-1) roots can be solved, which are around the outside and inside of the unit circle, the complex value root closest to the unit circle is closest to the real angle, and the ideal complex value root without noise and interference is located on the unit circle (complex amplitude is 1).

[0143] Step S145: Based on the K first target complex roots and the receiving element spacing of the frequency control array radar, the angle estimation value of each target is estimated. Specifically, the angle estimation value of each target is estimated according to the following formula :

[0144] (25).

[0145] Where, represents the phase angle of the complex number.

[0146] In some embodiments, step S150, based on the angle estimation value of each target, the polynomial root method is used to solve the multi-target parameter dimension reduction estimation model to estimate the distance estimation value of each target corresponding to the angle estimation value of each target, including:

[0147] Step S151: Substitute the angle estimation value of each target into the multi-target parameter dimension reduction estimation model respectively. Specifically, the angle estimation value of each target estimated in step S145 is substituted into the multi-target parameter dimension reduction estimation model, that is, substituted into the above formula (17).

[0148] Step S152: For any angle estimation value of a target, the second complex root of the polynomial when the corresponding multi-target parameter dimension reduction estimation model is 0 is solved. That is, for the kth target angle estimation value, the polynomial when the corresponding multi-target parameter dimension reduction estimation model is 0 is:

[0149] (26).

[0150] Rewrite formula (26) to get the polynomial about v as follows:

[0151] (27).

[0152] Where, .

[0153] Solve formula (27) by using the polynomial root method to get the corresponding second complex root v k . For each target angle estimation value, K second complex roots of K corresponding polynomials , k=1,2,…,K. ​

[0154] Step S153: Based on the second complex root and the frequency offset of the frequency-controlled array radar, estimate the distance of each target. Specifically, the distance of each target is estimated according to the following formula: :

[0155] (28).

[0156] In this embodiment, the distance of the target is estimated based on the estimated angle value of the target, and the estimation method of two-dimensional joint distance and angle estimation is no longer used, thereby reducing the calculation complexity.

[0157] Through experimental comparison, as shown in Table 1 below, the methods of the above embodiments greatly reduce the computational complexity compared with the two-dimensional multiple signal classification (2D-MUSIC) algorithm. Compared with the reduced-dimension multiple signal classification (RD-MUSIC) algorithm and the reduced-dimension multiple signal classification based on high-order singular value decomposition (HOSVD-RD-MUSIC) algorithm, that is, Figure 3-Figure 9 The tensor-based dimensionality reduction multiple signal classification algorithm in slightly reduces the computational complexity. In Table 1, represents the number of grid divisions for angle domain search, represents the number of grid divisions for distance domain search, and It is several orders of magnitude larger than M or N. The computational complexity expression of the algorithm in this embodiment does not include and ,Therefore, the computational complexity of the algorithm in this embodiment is the lowest.

[0158] Table 1 Comparison of algorithm theoretical computational complexity

[0159]

[0160] like Figure 3 As shown, assuming that the number of transmitting array elements is equal to the number of receiving array elements, Figure 3 A comparison chart showing the theoretical computational complexity of each algorithm as the number of array elements changes is shown. It can be seen that the algorithm of this embodiment has the lowest theoretical computational complexity.

[0161] Assuming the number of transmitting elementsM = 8, number of receiving elements N = 8, carrier center frequency = 3GHz, frequency offset = 20kHz, 100 Monte Carlo times for each experiment.

[0162] Under the above experimental conditions, as shown in Figure 4 and Figure 5 , assuming that the number of snapshots L = 6, three targets are located at (1km, -30°), (3.5km, 0°) and (2km, 45°), Figure 4 shows the change of the root mean square error (RMSE) of the estimated angle estimation value and the angle true value with the change of the signal-to-noise ratio (SNR), Figure 5 shows the change of the root mean square error of the distance estimation value and the distance true value with the change of the signal-to-noise ratio. It can be seen that the algorithm of the embodiment has the lowest root mean square error (the lower the root mean square error, the higher the estimation accuracy) compared with the existing dimension reduction estimation algorithm. Therefore, the algorithm is superior to the traditional two-dimensional spectrum peak search algorithm in performance under the condition of low snapshot number. Therefore, the algorithm of the embodiment can realize effective distance and angle estimation.

[0163] Under the above experimental conditions, as shown in Figure 6 and Figure 7 , assuming that the number of snapshots L = 6, two targets are located at (1km, -5°) and (3.5km, 0°), and the angles of the two targets are close (only differ by 5°). Figure 6 shows the change of the root mean square error of the angle estimation value and the angle true value with the change of the signal-to-noise ratio when the target angles are close, Figure 7 shows the change of the root mean square error of the distance estimation value and the distance true value with the change of the signal-to-noise ratio when the target angles are close. It can be observed that under the conditions of low snapshot number and low signal-to-noise ratio (SNR is below 12), when the target angles are close, the performance of the existing several dimension reduction estimation algorithms in the angle and distance estimation stage has sharply decreased, and the target loss occurs. In contrast, the algorithm proposed in the embodiment can still maintain high estimation accuracy, and has basically the same accuracy compared with the non-dimension reduction two-dimensional parameter estimation method.

[0164] Under the above experimental conditions, as shown in Figure 8 and Figure 9 , assuming that the signal-to-noise ratio SNR = 2dB, three targets are located at (1km, -10°), (3.5km, 0°) and (2km, 15°), and the angles of the three targets are close. Figure 8It shows how the root mean square error between the angle estimate and the true angle value changes with the number of snapshots when the angles of multiple targets are similar. Figure 9 This example shows how the root mean square error (RMS) between the estimated and true distances varies with the number of snapshots when multiple targets are close in angle. The algorithm proposed in this example achieves the lowest RMS error, while other existing dimensionality reduction estimation algorithms significantly degrade in performance under low snapshot conditions, leading to target loss.

[0165] The multi-target parameter dimensionality reduction estimation device for a frequency-controlled array radar provided by the present invention is described below. The multi-target parameter dimensionality reduction estimation device for a frequency-controlled array radar described below and the multi-target parameter dimensionality reduction estimation method for a frequency-controlled array radar described above can be referenced to each other.

[0166] The multi-target parameter dimensionality reduction estimation device for frequency-controlled array radar according to the embodiment of the present invention is as follows: Figure 10 Shown, including:

[0167] The received signal matrix determination module 1010 is used to determine the received signal matrix of the frequency-steering array radar based on the transmit steering vector and receive steering vector of the frequency-steering array radar for each target and the transmit signal of the frequency-steering array radar at each snapshot time.

[0168] The received signal matrix solving module 1020 is configured to solve the received signal matrix using a high-order singular value decomposition method to obtain a noise subspace matrix of the received signal matrix.

[0169] The model construction module 1030 is used to construct a multi-target parameter dimensionality reduction estimation model based on the noise subspace matrix and the transmission steering vector and the reception steering vector of the frequency-controlled array radar for each target.

[0170] The angle estimation module 1040 is used to solve the multi-objective parameter dimensionality reduction estimation model using a polynomial root-finding method to estimate the angle estimation value of each target.

[0171] The distance estimation module 1050 is used to solve the multi-target parameter dimensionality reduction estimation model based on the angle estimation value of each target using a polynomial root-finding method to estimate the distance estimation value of each target corresponding to the angle estimation value of each target.

[0172] In some embodiments, the received signal matrix determination module 1010 is specifically configured to:

[0173] Based on the transmitting and receiving steering vectors of the frequency-controlled array radar to each target, the joint transmitting and receiving array manifold matrix is ​​determined.

[0174] Based on the transmit signals at each snapshot moment, a preset noise vector, and the joint transmit-receive array manifold matrix, a receive signal matrix of the frequency-controlled array radar is determined, wherein the transmit signals at each snapshot moment all satisfy an orthogonality condition.

[0175] In some embodiments, the received signal matrix solving module 1020 is specifically configured to:

[0176] The received signal matrix is ​​reconstructed into a third-order tensor.

[0177] A high-order singular value decomposition is performed on the third-order tensor, and a truncation process is performed on the decomposed third-order tensor to obtain a signal subspace matrix based on the third-order tensor.

[0178] The signal subspace matrix is ​​sequentially subjected to normalization and orthogonal projection to obtain the noise subspace matrix.

[0179] In some embodiments, the model building module 1030 is specifically configured to:

[0180] The total steering vector is determined based on the transmitting steering vector and receiving steering vector of the frequency-controlled array radar for each target.

[0181] A cost function is constructed based on the total steering vector and the noise subspace matrix. The cost function includes a distance component in the transmit steering vector and a sign matrix. The sign matrix is ​​determined by an angle component in the transmit steering vector and a receive steering vector.

[0182] A Lagrangian function is constructed based on the cost function, wherein a Lagrangian multiplier in the Lagrangian function is determined based on when a partial derivative of the Lagrangian function on a distance portion of a transmission steering vector is equal to 0.

[0183] An estimate of the range portion of the transmit steering vector is determined based on the Lagrangian function.

[0184] The estimated value of the distance part in the transmission steering vector is substituted for the distance part in the transmission steering vector in the cost function to obtain a replaced cost function.

[0185] The minimized replaced cost function is determined to be the multi-objective parameter dimensionality reduction estimation model.

[0186] In some embodiments, the angle estimation module 1040 is specifically configured to:

[0187] When the multi-objective parameter dimensionality reduction estimation model reaches a minimum, the determinant of the symbol matrix is ​​set to 0.

[0188] Fast Fourier transform is used to calculate the determinant of the symbol matrix. When the determinant of the symbol matrix is ​​0, a polynomial of order 2M(N-1) is obtained, where M represents the number of transmitting elements of the frequency-controlled array radar and N represents the number of receiving elements of the frequency-controlled array radar.

[0189] Roots are obtained for the 2M(N-1)-order polynomial to obtain a plurality of first complex-valued roots.

[0190] K first target complex-valued roots closest to the unit circle are selected from multiple first complex-valued roots, where K represents the number of targets.

[0191] An angle estimation value of each target is estimated based on the K complex-valued roots of the first targets and the receiving array element spacing of the frequency-controlled array radar.

[0192] In some embodiments, the distance estimation module 1050 is specifically configured to:

[0193] The angle estimation value of each target is respectively substituted into the multi-target parameter dimensionality reduction estimation model.

[0194] For the angle estimation value of any target, solve the second complex root of the polynomial when the corresponding multi-target parameter dimensionality reduction estimation model is 0.

[0195] An estimated distance value of each target is estimated based on the second complex-valued root and the frequency offset of the frequency-controlled array radar.

[0196] Figure 11 An example of a physical structure diagram of an electronic device is shown below. Figure 11 As shown, the electronic device may include: a processor 1110, a communication interface 1120, a memory 1130, and a communication bus 1140, wherein the processor 1110, the communication interface 1120, and the memory 1130 communicate with each other via the communication bus 1140. The processor 1110 may call logic instructions in the memory 1130 to execute a multi-target parameter dimensionality reduction estimation method for a frequency-controlled array radar, the method comprising:

[0197] The receiving signal matrix of the frequency-steering array radar is determined based on the transmitting steering vector and receiving steering vector of the frequency-steering array radar to each target and the transmitting signal of the frequency-steering array radar at each snapshot moment.

[0198] A high-order singular value decomposition method is used to solve the received signal matrix to obtain a noise subspace matrix of the received signal matrix.

[0199] Based on the noise subspace matrix and the transmission steering vector and the reception steering vector of the frequency-controlled array radar for each target, a multi-target parameter dimensionality reduction estimation model is constructed.

[0200] The multi-objective parameter dimensionality reduction estimation model is solved by using a polynomial root-finding method to estimate the angle estimation value of each target.

[0201] Based on the angle estimation value of each target, the multi-target parameter dimensionality reduction estimation model is solved by using a polynomial rooting method to estimate the distance estimation value of each target corresponding to the angle estimation value of each target.

[0202] Furthermore, the logic instructions in the aforementioned memory 1130 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the portion that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product, stored in a storage medium, includes instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of the methods described in various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as a USB flash drive, a mobile hard drive, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.

[0203] On the other hand, the present invention further provides a computer program product, comprising a computer program. The computer program may be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer is capable of executing the multi-target parameter dimensionality reduction estimation method for a frequency-controlled array radar provided by the above methods. The method comprises:

[0204] The receiving signal matrix of the frequency-steering array radar is determined based on the transmitting steering vector and receiving steering vector of the frequency-steering array radar to each target and the transmitting signal of the frequency-steering array radar at each snapshot moment.

[0205] A high-order singular value decomposition method is used to solve the received signal matrix to obtain a noise subspace matrix of the received signal matrix.

[0206] Based on the noise subspace matrix and the transmission steering vector and the reception steering vector of the frequency-controlled array radar for each target, a multi-target parameter dimensionality reduction estimation model is constructed.

[0207] The multi-objective parameter dimensionality reduction estimation model is solved by using a polynomial root-finding method to estimate the angle estimation value of each target.

[0208] Based on the angle estimation value of each target, the multi-target parameter dimensionality reduction estimation model is solved by using a polynomial rooting method to estimate the distance estimation value of each target corresponding to the angle estimation value of each target.

[0209] In another aspect, the present invention further provides a non-transitory computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, the method for multi-target parameter dimensionality reduction estimation for a frequency-controlled array radar provided by the above methods is implemented. The method includes:

[0210] The receiving signal matrix of the frequency-steering array radar is determined based on the transmitting steering vector and receiving steering vector of the frequency-steering array radar to each target and the transmitting signal of the frequency-steering array radar at each snapshot moment.

[0211] A high-order singular value decomposition method is used to solve the received signal matrix to obtain a noise subspace matrix of the received signal matrix.

[0212] Based on the noise subspace matrix and the transmission steering vector and the reception steering vector of the frequency-controlled array radar for each target, a multi-target parameter dimensionality reduction estimation model is constructed.

[0213] The multi-objective parameter dimensionality reduction estimation model is solved by using a polynomial root-finding method to estimate the angle estimation value of each target.

[0214] Based on the angle estimation value of each target, the multi-target parameter dimensionality reduction estimation model is solved by using a polynomial rooting method to estimate the distance estimation value of each target corresponding to the angle estimation value of each target.

[0215] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, i.e., they may be located in one location or distributed across multiple network units. Some or all of the modules may be selected based on actual needs to achieve the objectives of the present embodiment. Persons of ordinary skill in the art will be able to understand and implement the present invention without inventive effort.

[0216] Through the above description of the embodiments, those skilled in the art will clearly understand that each embodiment can be implemented using software plus a necessary general-purpose hardware platform, or of course, hardware. Based on this understanding, the essence of the above technical solution, or the portion that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a magnetic disk, or an optical disk, and includes a number of instructions for causing a computer device (such as a personal computer, server, or network device) to execute the methods described in each embodiment or certain portions of the embodiments.

[0217] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the various embodiments of the present invention.

Claims

1. A multi-target parameter dimensionality reduction estimation method for frequency-controlled array radar, characterized in that: include: Determine the receiving signal matrix of the frequency-steering array radar based on the transmitting steering vector and receiving steering vector of the frequency-steering array radar for each target and the transmitting signal of the frequency-steering array radar at each snapshot moment; Solving the received signal matrix using a high-order singular value decomposition method to obtain a noise subspace matrix of the received signal matrix; Based on the noise subspace matrix and the transmission steering vector and the reception steering vector of the frequency-controlled array radar for each target, a multi-target parameter dimensionality reduction estimation model is constructed; A polynomial root-finding method is used to solve the multi-objective parameter dimensionality reduction estimation model to estimate the angle estimation value of each target; Based on the angle estimation value of each target, a polynomial rooting method is used to solve the multi-target parameter dimensionality reduction estimation model to estimate the distance estimation value of each target corresponding to the angle estimation value of each target; The high-order singular value decomposition method is used to solve the received signal matrix to obtain the noise subspace matrix of the received signal matrix, including: Reconstructing the received signal matrix into a third-order tensor; performing high-order singular value decomposition on the third-order tensor and performing truncation processing on the decomposed third-order tensor to obtain a signal subspace matrix based on the third-order tensor; performing a normalization operation and an orthogonal projection on the signal subspace matrix in sequence to obtain the noise subspace matrix; Wherein, based on the noise subspace matrix and the transmission steering vector and reception steering vector of the frequency-controlled array radar for each target, a multi-target parameter dimensionality reduction estimation model is constructed, including: Determine the total steering vector based on the transmit steering vector and receive steering vector of each target by the frequency-controlled array radar; constructing a cost function based on the total steering vector and the noise subspace matrix, the cost function including a distance portion of the transmit steering vector and a sign matrix, the sign matrix being determined by an angle portion of the transmit steering vector and a receive steering vector; Constructing a Lagrangian function based on the cost function, wherein a Lagrangian multiplier in the Lagrangian function is determined based on a partial derivative of the Lagrangian function with respect to a distance portion of a launch steering vector being equal to 0; determining an estimate of a range portion of a launch steering vector based on the Lagrangian function; replacing the distance portion of the transmit steering vector in the cost function with the estimated value of the distance portion in the transmit steering vector to obtain a replaced cost function; The minimized replaced cost function is determined to be the multi-objective parameter dimensionality reduction estimation model.

2. The multi-target parameter dimensionality reduction estimation method for frequency-controlled array radar according to claim 1, characterized in that: Based on the transmit steering vector and receive steering vector of the frequency-steering array radar for each target and the transmit signal of the frequency-steering array radar at each snapshot, the receive signal matrix of the frequency-steering array radar is determined, including: Based on the transmit and receive steering vectors of the frequency-controlled array radar to each target, the joint transmit and receive array manifold matrix is ​​determined; Based on the transmit signals at each snapshot moment, a preset noise vector, and the joint transmit-receive array manifold matrix, a receive signal matrix of the frequency-controlled array radar is determined, wherein the transmit signals at each snapshot moment all satisfy an orthogonality condition.

3. The multi-target parameter dimensionality reduction estimation method for frequency-controlled array radar according to claim 1, characterized in that: The multi-objective parameter dimensionality reduction estimation model is solved by using a polynomial rooting method to estimate the angle estimation value of each target, including: When the multi-objective parameter dimensionality reduction estimation model reaches a minimum, the determinant of the symbol matrix is ​​set to 0; The determinant of the symbol matrix is ​​calculated using fast Fourier transform. When the determinant of the symbol matrix is ​​0, a polynomial of order 2M(N-1) is obtained, where M represents the number of transmitting elements of the frequency-controlled array radar and N represents the number of receiving elements of the frequency-controlled array radar. Finding roots for the 2M(N-1)-order polynomial to obtain a plurality of first complex-valued roots; Select K first target complex-valued roots closest to the unit circle from multiple first complex-valued roots, where K represents the number of targets; An angle estimation value of each target is estimated based on the K complex-valued roots of the first targets and the receiving array element spacing of the frequency-controlled array radar.

4. The multi-target parameter dimensionality reduction estimation method for frequency-controlled array radar according to claim 1, characterized in that: Based on the angle estimation value of each target, a polynomial rooting method is used to solve the multi-target parameter dimensionality reduction estimation model to estimate the distance estimation value of each target corresponding to the angle estimation value of each target, including: Substituting the angle estimation value of each target into the multi-target parameter dimensionality reduction estimation model respectively; For the angle estimation value of any target, solve the second complex root of the polynomial when the corresponding multi-target parameter dimensionality reduction estimation model is 0; An estimated distance value of each target is estimated based on the second complex-valued root and the frequency offset of the frequency-controlled array radar.

5. A multi-target parameter dimensionality reduction estimation device for frequency-controlled array radar, characterized in that: include: a receiving signal matrix determination module, configured to determine the receiving signal matrix of the frequency-steering array radar based on the transmitting steering vector and receiving steering vector of the frequency-steering array radar for each target and the transmitting signal of the frequency-steering array radar at each snapshot; A receiving signal matrix solving module is used to solve the receiving signal matrix using a high-order singular value decomposition method to obtain a noise subspace matrix of the receiving signal matrix; A model building module is used to build a multi-target parameter dimensionality reduction estimation model based on the noise subspace matrix and the transmission steering vector and reception steering vector of the frequency-controlled array radar for each target; An angle estimation module, used to solve the multi-objective parameter dimensionality reduction estimation model using a polynomial root-finding method to estimate the angle estimation value of each target; A distance estimation module is used to solve the multi-target parameter dimensionality reduction estimation model based on the angle estimation value of each target using a polynomial root-finding method to estimate the distance estimation value of each target corresponding to the angle estimation value of each target; The received signal matrix solving module is specifically used for: Reconstructing the received signal matrix into a third-order tensor; performing high-order singular value decomposition on the third-order tensor and performing truncation processing on the decomposed third-order tensor to obtain a signal subspace matrix based on the third-order tensor; performing a normalization operation and an orthogonal projection on the signal subspace matrix in sequence to obtain the noise subspace matrix; The model building module is specifically used to: Determine the total steering vector based on the transmit steering vector and receive steering vector of each target by the frequency-controlled array radar; constructing a cost function based on the total steering vector and the noise subspace matrix, the cost function including a distance portion of the transmit steering vector and a sign matrix, the sign matrix being determined by an angle portion of the transmit steering vector and a receive steering vector; Constructing a Lagrangian function based on the cost function, wherein a Lagrangian multiplier in the Lagrangian function is determined based on a partial derivative of the Lagrangian function with respect to a distance portion of a launch steering vector being equal to 0; determining an estimate of a range portion of a launch steering vector based on the Lagrangian function; replacing the distance portion of the transmit steering vector in the cost function with the estimated value of the distance portion in the transmit steering vector to obtain a replaced cost function; The minimized replaced cost function is determined to be the multi-objective parameter dimensionality reduction estimation model.

6. An electronic device comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that: When the processor executes the computer program, the multi-target parameter dimensionality reduction estimation method for a frequency-controlled array radar is implemented as claimed in any one of claims 1 to 4.

7. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the multi-target parameter dimensionality reduction estimation method for a frequency-controlled array radar according to any one of claims 1 to 4 is implemented.

8. A computer program product comprising a computer program, characterized in that When the computer program is executed by a processor, the multi-target parameter dimensionality reduction estimation method for a frequency-controlled array radar according to any one of claims 1 to 4 is implemented.

Images