Multi-view high-resolution MIMO radar correlation imaging method and device based on AM algorithm

Abstract

This invention discloses a multi-view high-resolution MIMO radar correlation imaging method and apparatus based on the AM algorithm. The method simultaneously considers the influence of transceiver element position errors, equating them to transceiver phase errors. Drawing on the alternating iterative approach of the AM algorithm, it jointly estimates the transceiver element position errors and the target scattering coefficients, updating the reference signal using the estimated element position errors. Through continuous iteration, it corrects the element position errors, obtaining high-resolution multi-view MIMO radar correlation imaging results. This invention can correct transceiver element position errors and obtain high-resolution multi-view integrated imaging results.

Description

Technical Field

[0001] This invention belongs to the field of radar correlation imaging signal processing, specifically relating to a multi-view high-resolution MIMO radar correlation imaging method and device based on AM algorithm. Background Technology

[0002] Radar correlation imaging, a real-aperture radar imaging method inspired by optical ghost imaging, relies on the spatiotemporally uncorrelated transmitted signals to form a spatiotemporally random radiation field in the imaging region, and utilizes the spatiotemporal differences of the radiation field for imaging. Compared with traditional real-aperture radar imaging, its imaging resolution is related to the randomness of the radiation field, breaking through the limitation of antenna aperture on resolution and achieving high resolution capability. It also overcomes the shortcomings of synthetic aperture radar imaging, which is dependent on and constrained by motion. Therefore, radar correlation imaging is a high-resolution radar imaging technology that does not depend on the relative motion between the target and the radar, providing a new approach and direction for solving the shortcomings of traditional radar imaging, and has broad development prospects.

[0003] MIMO (multiple input multiple output) radar correlation imaging can improve imaging resolution by adding receiving channels, based on multi-input single-output radar correlation imaging. On the one hand, it expands the spatial dimension and increases the randomness of the radiation field reference signal, thereby improving the imaging resolution. On the other hand, it expands the observation perspective, which is beneficial for more comprehensively representing complex targets.

[0004] Radar correlation imaging requires accurate extrapolation of the radiation field reference signal based on the positions of the transceiver array elements, the transmitted signal, and the imaging unit. At this point, the extrapolated reference signal and the echo are perfectly matched. However, in reality, due to radar system errors (such as transceiver array asynchrony, element coupling, amplitude and phase errors, and position errors) and grid mismatch errors (such as relative motion between the target and the carrier platform, and the target scattering point not being at the center of the imaging unit), the extrapolation of the reference signal becomes incorrect, the correlation between the reference signal and the echo degrades, and model mismatch occurs, affecting the quality of correlation imaging. This invention mainly addresses the impact of MIMO radar transceiver array position errors on correlation imaging and proposes a multi-view MIMO radar correlation imaging position error correction method based on the AM algorithm.

[0005] Many scholars have conducted research on the mismatch problem of radar correlation imaging models and achieved good results. Existing research has proposed two array position error correction algorithms under a sparse Bayesian learning framework: one treats the position error of the transmitting elements as equivalent to a phase error, and the other constructs an imaging model under the array position error through Taylor expansion, assigning Gaussian-gamma-gamma priors to the target and correcting the position error through iterative iteration. Other methods use Taylor expansion to transform the nonlinear relationship between the echo and the element position error into a linear relationship, using auxiliary object correction, iterative compensation, and equivalent compensation methods to correct the position error, thus achieving array position correction and obtaining high-resolution correlation imaging results. However, existing research is focused on MISO (multiple input single output) correlation imaging, only considering the impact of the transmitting element position error on imaging.

[0006] This invention addresses the shortcomings of existing research by investigating a method for correcting the position error of a multi-view MIMO radar-based correlated imaging array. It considers the influence of transceiver element position errors, equating them to transceiver phase errors, and employs the iterative approach of the Alternating Minimization (AM) algorithm to jointly estimate the transceiver array position errors and the target scattering coefficients. Simulation results demonstrate that the proposed method can correct transceiver element position errors, yielding high-resolution multi-view integrated imaging results. Summary of the Invention

[0007] Objective: To address the problem of image quality degradation caused by transceiver array position errors in MIMO radar, this invention proposes a multi-view high-resolution MIMO radar correlation imaging method and device based on the AM algorithm. This method simultaneously considers the impact of transceiver array element position errors on image quality, jointly estimates the transceiver array element position errors and target scattering coefficients, and updates the reference signal using the estimated array element position errors. Through continuous iteration, the array element position errors are corrected, resulting in high-resolution multi-view MIMO radar correlation imaging results.

[0008] Technical solution: This invention provides a multi-view high-resolution MIMO radar correlation imaging method based on the AM algorithm, specifically including the following steps:

[0009] Step 1: The MIMO radar has M transmitting antennas and N receiving antennas. A set of independent noise amplitude-modulated signals is transmitted at the transmitting antennas, dividing the imaging plane into K = x * y imaging units, where x represents the number of lateral imaging units and y represents the number of longitudinal imaging units. The number of time samples is J. Based on the positions of the transmitting and receiving array elements and the imaging unit positions, N reference matrices A are obtained. n By concatenating N reference matrices along the time dimension, an extended-dimensional reference matrix is ​​obtained.

[0010] Step 2: Sample the signal at the receiving end for time J to obtain N echo vectors. After splicing, the extended-dimensional echo vector is obtained. The position error of the transceiver array elements is equated to the phase error of the transceiver array. A multi-view MIMO radar correlation imaging signal model is constructed, with the iteration number set to i=0 and the initial target scattering sparse vector... Transmit / receive phase error θ 0 =0;

[0011] Step 3: Based on the sparse reconstruction multi-view MIMO radar correlation imaging algorithm, the extended dimension reference matrix and the extended dimension echo are jointly processed to obtain the mean scattering coefficient of the target from multiple views. Preliminary estimate;

[0012] Step 4: Combining the AM alternating iterative approach, based on... Using prior information, combined with the extended-dimensional reference matrix and the extended-dimensional echo vector, the transmitted phase error is respectively... and the received phase error θ i+1 Make an estimate;

[0013] Step 5: Update the extended-dimensional reference matrix using the estimated transmit phase error and receive phase error, as shown below.

[0014] Step 6: Set the maximum number of iterations I max Given a suitable convergence threshold η, let i = i + 1, and determine whether the number of iterations has been reached or the convergence condition has been met. If the condition is not met, return to step 3; if it is met, the mean value of the multi-view target scattering coefficient after position error correction is obtained.

[0015] Step 7: Average the obtained multi-view target scattering coefficients The actual position of the grid is converted into an "x×y" matrix and displayed using MATLAB.

[0016] Furthermore, step (2) is achieved through the following formula:

[0017]

[0018]

[0019] in, For actual reference signal, For echo signal, S m (t) represents the noise amplitude-modulated signal generated by the m-th transmitting element. The path delay from the m-th transmitting element to the k-th imaging element and back to the n-th receiving element is given by [the path delay]. and θ n Let σ represent the phase error of the m-th transmitting element and the n-th receiving element, respectively. k,n It represents the scattering coefficient of the k-th imaging unit from the perspective of the n-th receiving element;

[0020] The multi-view MIMO radar correlated imaging signal model with element position errors is represented as follows:

[0021]

[0022] in, This represents the reference matrix under the nth receiving element when there is an element position error, and... and θ=[θ1,θ2,…,θ N ] T related, This represents the extended-dimensional reference matrix when there are element position errors. For the extended reference matrix derived before correction, diag(A) is the block diagonal matrix of the reference matrix, w is the noise vector, and σ n This represents the target scattering coefficient vector from the nth receiving viewpoint. This represents the average target scattering coefficient across N receiving viewpoints. The target scattering coefficient matrix under the nth receiving viewpoint and The difference between them.

[0023] Furthermore, step (3) is achieved through the following formula:

[0024]

[0025] Where ||·||1 represents the l1 norm of the vector, ||·||2 represents the l2 norm of the vector, and ||·|| TV The total variation regularization term for a vector is represented by the mathematical symbol *st*, which indicates that the regularization is restricted to a given vector. 0≤γ≤1 is a constraint term for the multi-view RCS fluctuation energy.

[0026] Furthermore, step (4) is achieved through the following formula:

[0027]

[0028]

[0029] Since the problem of estimating the phase error of the pairwise elements is a nonlinear least squares problem, a cost function is defined.

[0030] Number of The quasi-Newton method can be used to solve this problem:

[0031]

[0032]

[0033] in, and Let represent the gradient of the cost function with respect to the transmit phase error and the gradient of the cost function with respect to the receive phase error, respectively. and Represent the cost function pairs respectively The Hessian matrix of θ.

[0034] Based on the same inventive concept, the present invention also provides a multi-view high-resolution MIMO radar correlation imaging device based on AM algorithm, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the computer program is loaded into the processor, it implements the above-mentioned multi-view high-resolution MIMO radar correlation imaging method based on AM algorithm.

[0035] Beneficial Effects: Compared with existing technologies, the present invention offers the following advantages: The present invention employs a MIMO radar correlation imaging system. Compared to the MISO correlation imaging system, by increasing the detection dimension, the randomness of the radiation field reference signal is greater than that under the MISO system, which is beneficial for improving imaging resolution. Furthermore, under multi-view transmission, the increased receiving viewpoint and expanded observation viewpoint allow for a more comprehensive representation of complex targets. The present invention uses a multi-view MIMO radar correlation imaging algorithm based on sparse reconstruction, which reduces the impact of target RCS fluctuations on imaging quality. Through a single sparse imaging process, the accurate acquisition of the mean RCS of targets from multiple views is achieved. The present invention simultaneously considers both transmitting element errors and receiving element position errors. By jointly estimating the phase errors of the transmitting and receiving elements and the target scattering coefficient, errors can be effectively corrected, resulting in high-resolution radar imaging results. Attached Figure Description

[0036] Figure 1 This is a flowchart of a multi-view high-resolution MIMO radar correlation imaging method based on the AM algorithm;

[0037] Figure 2 This is a schematic diagram of the MIMO radar correlation imaging principle of the present invention;

[0038] Figure 3 This invention presents an ideal imaging scenario diagram that reflects the average RCS of multiple receiving angles when the target RCS changes with the receiving angle during simulation experiments, with a mean of 15 dBsm and a variance of 30 dBsm.

[0039] Figure 4 The simulation results of this invention, with the position error uniformly distributed at (-0.001, 0.001)m and the average equivalent phase error of about 6°, after 100 Monte Carlo experiments, are shown in the figure. Among them, (a) is the uncorrected imaging result and (b) is the imaging result after using the method of this invention.

[0040] Figure 5 The simulation results of this invention, with the position error uniformly distributed at (-0.005, 0.005)m and the average equivalent phase error of approximately 29°, after 100 Monte Carlo experiments, are shown. Among them, (a) is the uncorrected imaging result and (b) is the imaging result after using the method of this invention. Detailed Implementation

[0041] The present invention will now be described in further detail with reference to the accompanying drawings.

[0042] This invention proposes a method for correcting the position error of multi-view correlated imaging array elements in MIMO radar based on the AM algorithm, such as... Figure 1 As shown.

[0043] Step 1: Assume that the MIMO radar has M transmitting antennas and N receiving antennas, and both the transmitting and receiving arrays are uniform linear arrays. Transmit a set of independent noise amplitude-modulated signals S at the transmitting antenna. m (t). The imaging plane is divided into K = x * y imaging units, and it is assumed that the target scattering center is located at the center of the imaging unit, where x represents the number of horizontal imaging units, y represents the number of vertical imaging units, and the number of time samples is J. According to equation (1), the reference signal A(t, r) of the k-th imaging unit under the n-th receiving element can be derived. k ), where r k This represents the position vector of the k-th imaging unit. This represents the path delay from the transmitting element to the corresponding imaging element and back to the receiving element. Through derivation, N (J×L) reference matrices A can be obtained. n By concatenating the N reference matrices along the time dimension, an extended-dimensional reference matrix of (NJ×L) is obtained. (Uncorrected)

[0044]

[0045] Step 2: When there is an array element position error, the transmit / receive array element position error is equivalent to the transmit / receive phase error, and the actual reference signal is used. and echo signal This can be represented by equations (2) and (3). After sampling at the receiving end for time J, N (J×1) echo vectors can be obtained. After splicing, we obtain an extended dimension echo vector of (NJ×1). Set the iteration count i = 0, θ 0 =0.

[0046]

[0047]

[0048] in, and θ n Let σ represent the phase error of the m-th transmitting element and the n-th receiving element, respectively. k,n It represents the scattering coefficient of the k-th imaging unit from the perspective of the n-th receiving element.

[0049] Therefore, the multi-view MIMO radar correlated imaging signal model with element position errors can be represented as:

[0050]

[0051] in, This represents the echo vector of the nth receiving element when there is an element position error. This represents the extended-dimensional echo vector when there are array element position errors. This represents the reference matrix under the nth receiving element when there is an element position error, and... and θ=[θ1,θ2,…,θ N ] T related, Let represent the extended-dimensional reference matrix when there are element position errors. For the extended reference matrix derived before correction, diag(A) is the block diagonal matrix of the reference matrix, w is the noise vector, which is assumed to be Gaussian white noise in this model, and σ n This represents the target scattering coefficient vector from the nth receiving viewpoint. This represents the average target scattering coefficient across N receiving viewpoints. The target scattering coefficient matrix under the nth receiving viewpoint and The difference between them.

[0052] Step 3: The schematic diagram of radar correlation imaging is as follows Figure 2 As shown, the echo signal can be considered as a linear representation of the reference signal of each imaging unit. By correlating the echo with the reference signal, the target can be recovered. However, due to the influence of the position error of the transceiver array elements, a correlation mismatch occurs between the derived reference signal and the echo, which seriously affects the imaging quality. This invention first utilizes a multi-view MIMO radar correlation imaging algorithm based on sparse reconstruction to jointly process the extended dimension reference matrix and the extended dimension echo to realize the multi-view target scattering coefficient. Preliminary estimate.

[0053]

[0054] Where ||·||1 represents the l1 norm of the vector, ||·||2 represents the l2 norm of the vector, and ||·|| TV Let represent the total variation regularization term for the vector, where μ is the regularization parameter, and the mathematical symbol st refers to the regularization term subject to . 0≤γ≤1 is a constraint term on the fluctuation energy of the multi-view RCS (Radar Cross Section). Since it cannot be obtained directly, it can be estimated based on the echo energy.

[0055] The multi-view RCS fluctuation characteristics of complex targets satisfy a log-normal distribution. The probability density function expression of the random variable σ of the complex target's RCS is shown below:

[0056]

[0057] Where σ0 represents the median of the random variable σ, and ρ represents the ratio of the mean to the median, which is a variable parameter. By adjusting the value of ρ, it can be equivalent to Sweling1, Sweling3, and chi-square distribution models, which can fit various types of targets. This invention averages the target RCS results from multiple receiving perspectives, such as... Figure 3 As shown, this illustrates the ideal imaging scenario for complex targets from multiple perspectives.

[0058] Step 4: Combining the AM alternating iterative approach, based on... Using prior information, combined with the extended-dimensional reference matrix and the extended-dimensional echo vector, the transmitted phase error is respectively... and the received phase error θ i+1 The estimates are shown in equations (7) and (8).

[0059]

[0060]

[0061] Since the problem of estimating the phase error of the pairwise elements is a nonlinear least squares problem, the cost function is defined as follows: The quasi-Newton method can be used to solve this problem:

[0062]

[0063]

[0064] in, and Let represent the gradient of the cost function with respect to the transmit phase error and the gradient of the cost function with respect to the receive phase error, respectively. The specific expressions are shown in equations (11) and (12). and Represent the cost function pairs respectively The Hessian matrix of θ is given by equations (13) and (14), where Im(·) represents the imaginary part of the complex number. It can be represented by the following formula (15), It can be represented by the following formula (16), s represents the residual vector. m,n The expression for is shown in equation (17).

[0065]

[0066]

[0067]

[0068]

[0069]

[0070]

[0071]

[0072]

[0073]

[0074] Step 5: Since the extended dimension reference signal is related to the phase error of the transmit and receive array elements, the extended dimension reference matrix can be updated using the estimated transmit and receive phase errors, expressed as follows: This gradually brings it closer to the actual extended-dimensional reference matrix.

[0075] Step 6: Set the maximum number of iterations I max Given a suitable convergence threshold η, let i = i + 1 and determine whether the number of iterations has been reached or the convergence condition has been met. If the condition is not met, return to step 3; if the condition is met, the multi-view target scattering coefficient vector after position error correction can be obtained.

[0076] Step 7: Calculate the obtained multi-view target scattering coefficient vector The actual positions of the grid are converted into an (x×y) matrix, and then displayed using MATLAB, as shown below. Figure 4 (b) and Figure 5 (b) shows the imaging results. The target scattering intensity of each imaging unit is displayed by the color of the unit. The closer the color is to white, the greater the scattering coefficient. The color bar on the right represents the scattering coefficient intensity value of the corresponding color. Figure 4 (a) and Figure 5 (a) The uncorrected imaging results were plotted when the mean position error of the array elements was 0.001m and 0.005m, respectively. By comparison, when the position error of the array elements is small, the target can be recovered in both cases, but the imaging quality of the corrected algorithm is better. When the error is large, the uncorrected algorithm can no longer recover the target, while the method of the present invention can accurately recover the target by correcting the position error of the array elements.

[0077] Based on the same inventive concept, the present invention also provides a multi-view high-resolution MIMO radar correlation imaging device based on AM algorithm, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the computer program is loaded into the processor, it implements the above-mentioned multi-view high-resolution MIMO radar correlation imaging method based on AM algorithm.

[0078] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A multi-aperture high-resolution MIMO radar correlation imaging method based on AM algorithm, characterized in that, Includes the following steps: Step 1: the number of MIMO radar transmitting antennas is M, the number of receiving antennas is N, a group of independent noise amplitude modulation signals are transmitted at the transmitting antenna end, the imaging plane is divided into K=x*y imaging units, x represents the number of transverse imaging units, y represents the number of longitudinal imaging units, and the number of time samples is J; according to the positions of the transmitting and receiving elements and the positions of the imaging units, N reference matrices are obtained The N reference matrices are spliced in the time dimension to obtain an extended reference matrix ; Step 2: Sample the signal at the receiving end for time J to obtain N echo vectors. After splicing, the extended-dimensional echo vector is obtained. The positional error of the transceiver array elements is equivalent to the phase error of the transceiver array. A multi-view MIMO radar correlation imaging signal model is constructed, and the number of iterations is set. Initial target scattering sparse vector Transmit and receive phase error , ; Step 3: Based on the sparse reconstruction multi-view MIMO radar correlation imaging algorithm, the extended dimension reference matrix and the extended dimension echo are jointly processed to obtain the mean scattering coefficient of the target from multiple views. Preliminary estimate; Step 4: Combining the AM alternating iterative approach, based on... Using prior information, combined with the extended-dimensional reference matrix and the extended-dimensional echo vector, the transmitted phase error is respectively... and receiving phase error Make an estimate; Step 5: Update the extended-dimensional reference matrix using the estimated transmit phase error and receive phase error, as shown below. ; Step 6: Set the maximum number of iterations and a suitable convergence threshold ,make And determine whether the number of iterations has been reached or the convergence condition has been met. If the condition is not met, return to step 3; if the condition is met, the mean value of the multi-view target scattering coefficient after position error correction is obtained. ; Step 7: Average the obtained multi-view target scattering coefficients The grid is converted into an x×y matrix based on its actual position and then displayed using MATLAB.

2. The multi-view high-resolution MIMO radar correlation imaging method based on AM algorithm according to claim 1, characterized in that, Step 2 is achieved through the following formula: （2） （3） in, For actual reference signal, For echo signal, This represents the noise amplitude-modulated signal generated by the m-th transmitting element. The path delay from the m-th transmitting element to the k-th imaging element and back to the n-th receiving element is given by [the path delay]. and Let represent the phase errors of the m-th transmitting element and the n-th receiving element, respectively. It represents the scattering coefficient of the k-th imaging unit from the perspective of the n-th receiving element; The multi-view MIMO radar correlated imaging signal model with element position errors is represented as follows: （4） in, This represents the echo vector of the nth receiving element when there is an element position error. This represents the extended-dimensional echo vector when there are array element position errors. This represents the reference matrix under the nth receiving element when there is an element position error, and... and related, This represents the extended-dimensional reference matrix when there are element position errors. Corresponding to the extended dimension reference matrix derived without correction, The block diagonal matrix of the reference matrix. For noise vectors, This represents the target scattering coefficient vector from the nth receiving viewpoint. This represents the average target scattering coefficient across N receiving viewpoints. The target scattering coefficient matrix under the nth receiving viewpoint and The difference between them.

3. The multi-view high-resolution MIMO radar correlation imaging method based on the AM algorithm according to claim 2, characterized in that, Step 3 is achieved through the following formula: （5） in, Representing vectors Norm, Representing vectors Norm, Represents the total variation regularization term for a vector. For regularization parameters, use mathematical symbols. This refers to being limited by, It is a constraint term on the energy fluctuations of the multi-view RCS.

4. The multi-view high-resolution MIMO radar correlation imaging method based on the AM algorithm according to claim 3, characterized in that, Step 4 is achieved through the following formula: （7） （8） Since the problem of estimating the phase error of the pairwise elements is a nonlinear least squares problem, the cost function is defined as follows: The quasi-Newton method can be used to solve this problem: （9） （10） in, and Let represent the gradient of the cost function with respect to the transmit phase error and the gradient of the cost function with respect to the receive phase error, respectively. and Represent the cost function pairs respectively and The Hessian matrix.

5. A multi-view high-resolution MIMO radar correlation imaging device based on AM algorithm, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the computer program is loaded into the processor, it implements the multi-view high-resolution MIMO radar correlation imaging method based on the AM algorithm according to any one of claims 1-4.

Images