Method and device for correcting amplitude-phase error of ground-based MIMO SAR, and storage medium

Abstract

The invention relates to the technical field of radar signal processing, in particular to a ground-based MIMO SAR amplitude-phase error correction method and device and a storage medium, and the method mainly comprises the steps: correcting an original echo signal of two-dimensional echo data according to a compensation matrix, and obtaining initial correction data; eliminating invalid distance units of the initial correction data based on a preset energy threshold value, selecting and sorting the remaining units according to the calculated signal-to-clutter ratio and normalized amplitude variance, eliminating linear phase offset of the distance units, aligning the strongest scattering point to the center, iteratively rotating the initial phases of the distance units to enable the directions of the initial phases to be consistent, and obtaining the initial correction data of the distance units. Obtaining first processing data; and estimating a channel phase error through a weighted least square method based on the unwrapped phase and the phase noise variance obtained by each iteration. According to the invention, the problems of poor adaptability to complex scenes and low correction precision in the prior art can be solved, and millimeter-level precision is realized.

Description

Technical Field

[0001] This application relates to the field of radar signal processing technology, and in particular to a method, apparatus and storage medium for correcting amplitude and phase errors of ground-based MIMO SAR. Background Technology

[0002] Traditional ground-based MIMO SAR systems have long faced severe challenges in channel amplitude and phase error correction. On the one hand, differences in antenna hardware, signal link disturbances, and phase center approximation errors in the equivalent virtual array lead to defocusing and sidelobe distortion in imaging. On the other hand, while existing single-point calibration methods can achieve initial correction using corner reflectors, they are ineffective against residual phase errors caused by array position deviations. When compensating for residual errors, traditional least squares algorithms ignore the dynamic differences in signal-to-clutter ratios (SCRs) across different range cells, treating high-noise data and strong target signals interchangeably. This results in clutter phase noise contaminating the estimation results in low SCR regions, leading to undercompensation or overcompensation.

[0003] In existing technologies, there are ways to improve the correction effect through phase expansion or iterative optimization, but these are always limited by two technical bottlenecks: First, a quantitative correlation between phase noise variance and signal-to-clutter ratio has not been established, making it impossible to distinguish the influence weight of isolated strong scattering points (such as houses) and distributed weak clutter (such as vegetation) on error estimation. Second, phase jump correction models with fixed thresholds are difficult to adapt to the complex and ever-changing monitoring environment of ground-based radar, and often fail, especially in strong interference scenarios such as bridge vibration monitoring and landslide early warning.

[0004] Therefore, there is an urgent need for a more refined method to correct the amplitude and phase errors of ground-based MIMO SAR, in order to solve the problem that traditional correction methods are not accurate in estimating residual phase errors in complex scenarios. Summary of the Invention

[0005] This application provides a method for correcting amplitude and phase errors of ground-based MIMO SAR, which can solve the problems of poor adaptability to complex scenes and low correction accuracy in the prior art, and achieve millimeter-level accuracy.

[0006] To achieve the above objectives, the present invention adopts the following technical solution: In a first aspect, the present invention provides a method for correcting amplitude and phase errors of ground-based MIMO SAR, comprising: Two-dimensional echo data is acquired from a corner reflector; The amplitude and phase errors of each radar channel relative to the reference channel are calculated based on the two-dimensional echo data, and a compensation matrix is ​​generated. The original echo signal of the two-dimensional echo data is corrected according to the compensation matrix to obtain the initial corrected data; Based on the preset energy threshold, invalid range cells in the initial correction data are removed. The remaining cells are selected and sorted according to the calculated signal-to-noise ratio and normalized amplitude variance. The linear phase shift of the range cells is eliminated and the strongest scattering point is aligned to the center. The initial phase of each range cell is iteratively rotated to make their directions consistent, and the first processed data is obtained. The distance units in the first processed data are divided according to a preset variance threshold to obtain high signal-to-noise ratio distance units and low signal-to-noise ratio distance units; The distance unit ranked first in the first processed data is used as the initial distance unit, and the process is iterated; During iteration, the phase of the fixed initial phase is expanded to obtain the expanded phase. Then, the high signal-to-noise ratio range cells and the low signal-to-noise ratio range cells are distinguished and calculated to obtain the corresponding phase noise variance. Based on the expanded phase and the phase noise variance obtained in each iteration, the channel phase error is estimated by weighted least squares method. The channel phase error is used to compensate for the error in the first processed data in each iteration; In a preferred example of this application, the preset variance threshold can be further set to 0.27.

[0007] In a preferred embodiment of this application, the step of dividing the distance units in the first processed data according to a preset variance threshold to obtain high signal-to-noise ratio distance units and low signal-to-noise ratio distance units includes: The distance cells in the first processed data whose normalized amplitude variance is less than a preset variance threshold are used as high signal-to-noise ratio distance cells. The distance cells in the first processed data whose normalized amplitude variance is greater than a preset variance threshold are used as low signal-to-noise ratio distance cells.

[0008] In a preferred embodiment of this application, the step of distinguishing and calculating the high signal-to-noise ratio (SNR) range cells and low SNR range cells to obtain their respective phase noise variances includes: For distance cells in the first processed data whose normalized amplitude variance is less than a preset variance threshold, the phase noise variance is calculated using a signal-to-noise ratio analytical model. The analytical model calculates the phase noise variance through the quantitative relationship of the signal-to-noise ratio. During iteration, for distance cells in the first processed data whose normalized amplitude variance is greater than a preset variance threshold, the phase noise variance is calculated using an iterative model based on residual error. The iterative model updates the current phase noise variance using the residual error estimate from the previous iteration step.

[0009] In a preferred embodiment of this application, the step of estimating the channel phase error using weighted least squares based on the expanded phase and the phase noise variance obtained in each iteration includes: A weight matrix is ​​constructed using the reciprocal of the phase noise variance as the weighting coefficient; Based on the weight matrix, the expanded phase is weighted and averaged to obtain the residual phase error.

[0010] In a preferred embodiment of this application, the step of using the channel phase error to compensate for errors in the first processed data of each iteration includes: The residual phase error is used to compensate for errors in the first processed data of each iteration.

[0011] Secondly, this application provides a ground-based MIMO SAR amplitude and phase error correction device, the device comprising: The data acquisition module is used to acquire two-dimensional echo data from the corner reflector; The initial correction module is used to calculate the amplitude and phase errors of each radar channel relative to the reference channel based on the two-dimensional echo data, and generate a compensation matrix; and to correct the original echo signal of the two-dimensional echo data according to the compensation matrix to obtain the initial correction data. The distance filtering module is used to remove invalid distance cells from the initial correction data based on a preset energy threshold, select and sort the remaining cells according to the calculated signal-to-noise ratio and normalized amplitude variance, eliminate the linear phase shift of the distance cells and align the strongest scattering point to the center, iteratively rotate the initial phase of each distance cell to make their directions consistent, and obtain the first processed data. An iterative module is used to divide the distance units in the first processed data according to a preset variance threshold to obtain high signal-to-noise ratio (SNR) distance units and low SNR distance units; the distance unit ranked first in the first processed data is used as the initial distance unit for iteration; during iteration, a phase expansion is performed on a fixed initial phase to obtain the expanded phase, and then the high SNR and low SNR distance units are distinguished and calculated to obtain the corresponding phase noise variance; based on the expanded phase and the phase noise variance obtained in each iteration, the channel phase error is estimated by weighted least squares method. The output module is used to obtain output data after all distance units in the first processed data have been iterated.

[0012] Thirdly, this application provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the ground-based MIMO SAR amplitude and phase error correction method as described in any of the preceding claims.

[0013] Fourthly, this application provides a computer-readable storage medium storing a program, wherein when the program is executed by a processor, it implements the ground-based MIMO SAR amplitude and phase error correction method as described in any of the preceding claims.

[0014] Fifthly, this application provides a computer program product including computer instructions that, when executed by a processor, implement the steps of the ground-based MIMO SAR amplitude and phase error correction method as described in any of the preceding claims.

[0015] In summary, compared with the prior art, the beneficial effects of the technical solution provided in this application include at least the following: 1. The corner reflector compensation in the initial calibration stage eliminates most of the inter-channel amplitude and phase errors, significantly improving the imaging defocus problem. Then, the fine calibration stage uses weighted least squares estimation to suppress the standard deviation of the residual phase error, solving the problems of poor adaptability to complex scenes and low calibration accuracy in existing technologies, and achieving millimeter-level accuracy.

[0016] 2. The data is divided into high and low signal-to-noise ratio regions based on a preset variance threshold. The variance is calculated using a targeted calculation method for different signal-to-noise ratio regions. This can adapt to complex scenarios such as bridge vibration and vegetation cover, and suppress the interference of clutter phase noise on weak targets.

[0017] 3. Linear phase shift is eliminated through iteration, and the random initial phase entanglement problem is solved by combining initial phase alignment. A weight matrix is ​​constructed using the inverse of the phase noise variance, and then deviation compensation is performed using the weight matrix, ensuring stable convergence under strong interference and greatly improving the accuracy of error compensation. Attached Figure Description

[0018] Figure 1 This is a flowchart illustrating a method for correcting amplitude and phase errors in ground-based MIMO SAR, as provided in one embodiment of this application.

[0019] Figure 2 This is a data flow diagram illustrating a ground-based MIMO SAR amplitude and phase error correction method provided in one embodiment of this application.

[0020] Figure 3 Phase correction diagram of range cell for a ground-based MIMO SAR amplitude and phase error correction method provided in one embodiment of this application.

[0021] Figure 4 This is a block diagram of a ground-based MIMO SAR amplitude and phase error correction device provided in one embodiment of this application. Detailed Implementation

[0022] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0023] In one embodiment of this application, a method for correcting amplitude and phase errors of ground-based MIMO SAR is provided. Please refer to [link to relevant documentation]. Figure 1 As shown, the method includes: S100: Acquire two-dimensional echo data from the corner reflector.

[0024] S200: Calculate the amplitude and phase errors of each radar channel relative to the reference channel based on the two-dimensional echo data, and generate a compensation matrix.

[0025] S300: Correct the original echo signal of the two-dimensional echo data according to the compensation matrix to obtain the initial correction data.

[0026] S400: Based on the preset energy threshold, invalid range cells in the initial correction data are removed. The remaining cells are selected and sorted according to the calculated signal-to-noise ratio and normalized amplitude variance. The linear phase shift of the range cells is eliminated and the strongest scattering point is aligned to the center. The initial phase of each range cell is iteratively rotated to make their directions consistent, and the first processed data is obtained. S500: Divide the distance units in the first processed data according to the preset variance threshold to obtain high signal-to-noise ratio distance units and low signal-to-noise ratio distance units; take the distance unit ranked first in the first processed data as the initial distance unit and perform iteration; during iteration, perform phase expansion on the fixed initial phase to obtain the expanded phase, and then distinguish and calculate the high signal-to-noise ratio distance units and low signal-to-noise ratio distance units to obtain the corresponding phase noise variance.

[0027] S600: Based on the expanded phase and the phase noise variance obtained in each iteration, the channel phase error is estimated by weighted least squares method; the channel phase error is used to compensate for the error of the first processed data in each iteration; when all distance cells in the first processed data have been iterated, the output data is obtained.

[0028] Specifically, the flowchart of steps S100-S600 is as follows: Figure 2 As shown.

[0029] In step S100, the corner reflector is a single-point strong scatterer with a known angle placed in the monitoring scene. It serves as the target for echo data observation and is used as a calibration source to acquire the two-dimensional echo data.

[0030] Amplitude and phase errors in the transmit and receive channels will cause amplitude and phase errors in the equivalent received virtual channel, resulting in deviations in the original echo. By performing multiple rounds of scene data acquisition and improving the signal-to-noise ratio, the baseband's original echo data can be used in a single... The matrix representation of , where Indicates the number of virtual channels (which can be upsampled). This represents the number of samples within a pulse. Channel amplitude and phase errors are superimposed on the virtual channel dimension, affecting imaging.

[0031] Considering the amplitude and phase errors on the virtual channel, we assume that the amplitude and phase errors caused by the actual transmit and receive channels result in the first... The amplitude and phase errors of each virtual channel are: ;(twenty two) In the presence of amplitude and phase errors, the range-compressed signal becomes: ;(twenty three) The purpose of amplitude and phase error correction is to solve or approximate the amplitude and phase errors of all virtual channels, and then compensate the distance-compressed signal to recover the original signal.

[0032] As described in steps S200 to S300, the echo data undergoes initial amplitude and phase error correction.

[0033] Firstly, from the perspective of the virtual array, the virtual array equivalent to the ground-based MIMO SAR radar transceiver antenna is a uniform linear array. Therefore, the amplitude and phase errors can be analyzed from the perspective of array signal processing. In the two-dimensional matrix representation, The corresponding representation is the number of virtual array elements. This represents the number of snapshots. According to the classic array signal model, the echo signal can be represented as: ;(twenty four) in, for vector representation The echo data received by each virtual array element at one time for vector representation The signal reflected by the target source For the number of information sources, for The vector representation of complex zero-mean additive white Gaussian noise, for The matrix, also known as the steering vector, is: (25) and This represents the amplitude and phase error matrix that introduces errors into each virtual channel. for The matrix is ​​defined as: (26) The key to amplitude and phase error correction is estimating the amplitude and phase error matrix. .

[0034] The specific correction process and algorithm derivation are as follows.

[0035] First, a single-point strong scattering object with a known angle is placed in the monitoring scene as the target for echo data observation (such as an angular reflector). This target is used as the calibration source. The signal after the radar transmitted signal is reflected by this source is... Then at this time the first... The signal received by each virtual channel can be represented as: (27) in The angle of the known calibration source.

[0036] Rewriting the above equation, we get: (28) To simplify the above formula, we define: (29) Then equation (28) can be written as: (30) During calibration, the first channel can usually be selected as a reference, then... , At this point, for the first channel: (31) Combining equations (30) and (31), we define: (32) Rewriting the above equation, we get: (33) Since the correction source and the noise are incoherent, and the noise is independent, the covariance matrix of the above equation can be calculated as follows: (34) Next, we will solve this problem. Covariance matrix: (35) Substituting equation (35) into equation (34), we get: (36) in This represents the signal-to-noise ratio (SNR) of the echo signal.

[0037] In amplitude and phase error estimation, assuming that both amplitude and phase errors are random, and given a fixed maximum allowable amplitude error offset and a high signal-to-noise ratio, equation (36) can be used to derive: (37) For example, in the case of the largest amplitude deviation, assuming And at this time ,but: (38) Therefore, based on the above derivation, we have: (39) In actual data processing, multiple sampling snapshots are generally required for calculation, so the final estimated value of the amplitude and phase error can be obtained as follows: (40) When using this method for correction in real-world scenarios, the calibration source is generally placed on the normal line of the radar monitoring area. At this time, the angle of the echo target is 0, and the echo signal-to-noise ratio is relatively high. Therefore, a relatively accurate correction result and amplitude-phase error estimate can be obtained. At this time, equation (40) becomes: (41) The above steps will yield the result. Initial values ​​for amplitude and phase error estimation for each virtual transceiver channel: (42) After initial correction, the basic amplitude error is negligible. However, due to the discrepancy between the actual radar transceiver array's position and the equivalent virtual array, some residual phase error still needs to be estimated and corrected. Phase estimation for the fine correction phase will be performed after imaging, using a weighted least squares algorithm. The specific algorithm principle is as follows.

[0038] After the echo signal data is compressed in the distance dimension, the 6th dimension is now... The phase of the echo signal of each distance cell can be expressed as: (43) in Indicates the first Doppler frequency shift of the strongest scattering point in each range cell The resulting phase, Indicates the first The initial phase existing in each distance cell, This represents phase noise caused by clutter. This indicates the phase error caused by inconsistencies between channels.

[0039] Fine-tuning occurs after the initial calibration, at which point the initial estimate of the channel phase error has been obtained through the initial calibration. Therefore, equation (43) can be rewritten as: (44) in This represents the residual channel phase error. In each phase part of the above equation, the first term is caused by the Doppler frequency shift of the strongest scattering point, which can be removed by cyclically shifting the strongest point of each range cell. The second term is a fixed phase term, which can be removed by phase alignment and expansion, which will be described in detail later. After these two steps, equation (44) can be simplified to: (45) The key issue is estimating the channel phase error from the phase data. It's typically assumed that clutter within each range cell is independent, thus causing phase noise. They are also independent of each other. Because... The estimation will introduce bias, so the objective is to minimize the variance of the estimation error, which becomes a weighted least squares (WLS) problem. The above equation can be rewritten in matrix form as follows: (46) in: (47) (48) (49) Phase noise without loss of generality It follows a zero-mean Gaussian distribution, and the th The variance of each distance cell is Then the weight matrix can be defined as follows: (50) The WLS estimate of the channel phase error is then: (51) Substituting equations (47) and (48) into equation (51), we obtain the following results: (52) From the above formula, we can see that the WLS estimate of the channel phase error is... It refers to the phase data in each range direction. The result of the weighted average, with the weights as follows: (53) It can be seen that the phase data weight of each distance cell is inversely proportional to the phase noise variance of that cell.

[0040] Therefore, by estimating the phase data of each distance unit and the phase data weight (or phase noise variance), the estimated value of the channel phase error can be calculated by equation (52).

[0041] First, we derive the formula for calculating the phase noise variance when the signal-to-clutter ratio (SCR) is large in the range cell. In this case, the range cell has a large SCR and is generally considered a prominent cell, meaning that the echo sequence of this range cell contains only isolated strong scattering centers with constant amplitude, while the other smaller scattering centers are considered clutter and noise.

[0042] Assume the first The strongest scatterer in each range cell has been shifted to the image center via cyclic shifting, while the initial phase Since the phase noise variance is assumed to be 0 and the inter-channel phase error is ignored, the signal of the strongest scattering point in this range cell can be simplified as follows: The remaining clutter signals are represented as ,in The complete error-free signal is represented as: (54) If we let: (55) (56) The phase noise can then be expressed as: (57) Expanding the above equation using Taylor series, we have: (58) because ,so From equation (58), we can see that: (59) Taking the average of the above equations and combining it with equations (55) and (56), we can obtain the following results: (60) Clearly, the above formula is related to the signal-to-noise ratio (SCR), which is: (61) Therefore, equation (60) can be further rewritten as: (62) In other words, for a range cell with a large signal-to-noise ratio, it is only necessary to first estimate the signal-to-noise ratio of the range cell, and then calculate the corresponding phase noise variance according to equation (62).

[0043] Since the phase of the echo signal is disrupted, the SCR of each range cell can only be estimated by amplitude. Starting from equation (54), we have: (63) Similarly, performing a Taylor expansion on the above equation, it can be simplified to: (64) Calculate the mean of the above equation and substitute it into equation (61) to obtain: (65) It is easy to see from equation (63): (66) Similarly, by calculating the mean of the above equation and substituting it into equation (61), we can obtain: (67) Replacing equations (65) and (66) with variables and calculating based on discrete data, we have: (68) (69) Combining equations (65), (67), (68), and (69), we can obtain the following expression: (70) Therefore, the mean and mean square values ​​of the amplitude of each range cell can be calculated first, and then the signal-to-noise ratio (SNR) of the corresponding range cell can be calculated according to equation (70). For range cells with a large SNR, the phase noise variance can be calculated according to equation (62) to obtain the weight. For range cells with a small SNR, the following method is needed to estimate the phase noise variance.

[0044] If the channel phase error Given that, then the first The phase noise variance of a distance cell can be expressed as: (71) In practical applications, phase error The value is unknown and generally changes iteratively; therefore, its estimated value can be used instead of the accurate value, and this error has a small impact when the signal-to-noise ratio is small. At this point, the... The phase noise variance of a distance cell can be expressed as: (72) Therefore, when calculating the phase noise variance of each distance cell, the SCR of each cell must first be calculated. Then, the phase noise variance of the distance cell with a large SCR is calculated by applying equation (62). Finally, the phase noise variance of the distance cell with a small SCR is calculated by applying equation (72) in the iteration.

[0045] To define the size of the SCR (Scattered Radiation Response), the prominent point cell selection method is chosen. First, the energy of each range cell is calculated, and range cells below a certain energy threshold are removed. These range cells have almost no scattering point echoes, only some very weak clutter and noise. Such range cells have no effect on the estimation and do not affect the estimation results. Then, the following formula is applied to the remaining range cells to calculate the variance of the normalized amplitude: (73) Generally speaking, select The distance cells are used as prominent point cells. In distinguishing between large and small signal-to-noise ratio distance cells, the distance cells with normalized amplitude variance less than 0.27 are selected as large signal-to-noise ratio cells, and the rest are small signal-to-noise ratio cells.

[0046] After determining the weights of each distance cell, it is also necessary to solve for the phase of each cell. As we know from the previous analysis, we first need to cyclically shift the strongest scattering point to the center of the image to remove the linear phase term. This step is relatively easy to implement. Then, we need to estimate and remove the fixed phase that exists in each distance cell.

[0047] For fixed phase terms present in each range cell, the randomness and inconsistency of the initial phase of each cell pose significant challenges to compensation. The existence of random initial phases causes the signals in each range cell to have different orientations in the complex plane. It is easy to conceive of first adjusting the random initial phase orientations of each range cell to the same direction, and then performing unified correction, such as... Figure 3 As shown, taking the first distance cell as a reference, the initial phases of all distance cells can be aligned by rotating the initial phases of the other distance cells. The sum of the 2-norm of the signals of all distance cells is maximized when they are in the same direction.

[0048] If we take the distance cell with the strongest scatterer intensity as a reference, the orientation alignment problem is transformed into the following function optimization problem: (74) in This indicates the order of the scatterer intensity. Echo data from each distance cell, Indicates the first The phase needs to be rotated when aligning each distance cell. After aligning all distance cells, a uniform fixed phase remains. Then, the phase can be compensated.

[0049] During phase alignment, the optimal distance unit is used first. For reference, then phase needs to be used. The second unit with suboptimal alignment The standard for alignment at this time is Then, the two are coherently superimposed to obtain Then, the process is iterated until all elements are aligned. During this process, the phase compensation for each element is... It is unknown, therefore it needs to be solved first. For example, expand the alignment standard: (75) in represent The phase, when When the above expression reaches its maximum value, at this time: (76) This represents taking the phase. Using the above equation as the initial condition, we can obtain... General estimation formula: (77) get After obtaining the estimated values, the signals are coherently superimposed using the following formula: ; (78) Finally, the alignment compensation phase of all range cells can be obtained by alternating iterative equations (77) and (78). The following initial phase compensation function is used to align all range cells: (79) After alignment, all distance cells now share the same fixed initial phase. If this item does not exist, then the phase will be as described above. It will be about channel phase error The central distribution, when this fixed phase exists, will cause the distribution to become random. Phase entanglement occurs, causing from Received The estimation is biased.

[0050] To compensate for this fixed initial phase, Local Phase Unwrapping (LPU) can be used. The principle is that if a phase signal is correctly unwrapped locally, it will match the channel phase error. The distance is closest. Clearly, if the initial phase... If the value of is known, then its expanded value must belong to the following set: (80) Selected phase It should belong to the set above, and with There is a closest distance, that is to say: (81) Due to the initial phase in actual conditions Since the value is unknown, after initial phase alignment is completed, a one-dimensional search can be performed during local phase unwrapping to find the initial phase with the minimum distance from the channel phase error. .

[0051] The distance can be calculated using the following formula, and the point with... The closest unfolding phase: (82) Obviously, It is about the first phase. The purpose of local phase expansion is to find a function that minimizes... First encounter and the corresponding unfolding phase During this process, the phase needs to be unfolded, and the channel phase error... Since the value is also unknown, an estimated value needs to be used instead. Therefore, an iterative method is needed to estimate and compensate sequentially.

[0052] The flowchart of the entire correction algorithm is shown below. After obtaining the two-dimensional original echo signal, the amplitude and phase errors are first calculated and compensated using the initial correction method, and then fine correction is performed. Fine correction occurs after range compression. The first step is to select and sort the range cells according to their signal-to-noise ratio (SNR) and normalized amplitude variance. Specifically, this involves: 1. Determining the energy threshold and eliminating range cells with energy below the threshold; 2. Calculating the normalized amplitude variance and SNR of each range cell and sorting them in descending order of SNR.

[0053] The second step is to shift the center of the strongest scattering point of each distance cell obtained in the first step and remove the linear phase term.

[0054] The third step is to perform initial alignment on the previously obtained distance cells.

[0055] Then, the first sorted distance cell is used as the initial condition for iteration. During the iteration, the phase is first expanded on the fixed initial phase to obtain the expanded phase. Then, the phase noise variance is calculated using equation (62) for distance cells with normalized amplitude variance less than 0.27, and the phase noise variance is calculated using equation (72) for distance cells with normalized amplitude variance greater than 0.27.

[0056] After obtaining the expanded phase and phase noise variance in each iteration, the channel phase error is estimated using equation (52). After all distance cells have been iterated, the final channel phase error estimate is used for compensation.

[0057] In this embodiment, corner reflector compensation during the initial correction stage eliminates most of the inter-channel amplitude and phase errors, significantly improving the imaging defocus problem. Then, in the fine correction stage, weighted least squares estimation is used to suppress the standard deviation of the residual phase error, solving the problems of poor adaptability to complex scenes and low correction accuracy in the prior art, and achieving millimeter-level accuracy.

[0058] The data is divided into high and low signal-to-noise ratio regions by a preset variance threshold. The variance is calculated using a targeted calculation method for different signal-to-noise ratio regions. This can adapt to complex scenarios such as bridge vibration and vegetation cover, and suppress the interference of clutter phase noise on weak targets.

[0059] Linear phase shift is eliminated iteratively, and the random initial phase entanglement problem is solved by combining initial phase alignment. A weight matrix is ​​constructed using the inverse of the phase noise variance, and then bias compensation is performed using this weight matrix, ensuring stable convergence under strong interference and greatly improving the accuracy of error compensation.

[0060] In some embodiments, the preset variance threshold is 0.27.

[0061] In some embodiments, dividing the distance units in the first processed data according to a preset variance threshold to obtain high signal-to-noise ratio distance units and low signal-to-noise ratio distance units includes: The distance cells in the first processed data whose normalized amplitude variance is less than a preset variance threshold are used as high signal-to-noise ratio distance cells. The distance cells in the first processed data whose normalized amplitude variance is greater than a preset variance threshold are used as low signal-to-noise ratio distance cells.

[0062] In some embodiments, the step of distinguishing and calculating the high signal-to-noise ratio range cells and low signal-to-noise ratio range cells to obtain their corresponding phase noise variances includes: For distance cells in the first processed data whose normalized amplitude variance is less than a preset variance threshold, the phase noise variance is calculated using a signal-to-noise ratio analytical model. The analytical model calculates the phase noise variance through the quantitative relationship of the signal-to-noise ratio. During iteration, for distance cells in the first processed data whose normalized amplitude variance is greater than a preset variance threshold, the phase noise variance is calculated using an iterative model based on residual error. The iterative model updates the current phase noise variance using the residual error estimate from the previous iteration step.

[0063] In some embodiments, estimating the channel phase error using weighted least squares based on the expanded phase and the phase noise variance obtained in each iteration includes: A weight matrix is ​​constructed using the reciprocal of the phase noise variance as the weighting coefficient; Based on the weight matrix, the expanded phase is weighted and averaged to obtain the residual phase error.

[0064] In some embodiments, the step of using the channel phase error to compensate for errors in the first processed data of each iteration includes: The residual phase error is used to compensate for errors in the first processed data of each iteration.

[0065] This application also provides a ground-based MIMO SAR amplitude and phase error correction device; please refer to [link to relevant documentation]. Figure 4 As shown, the device includes: Data acquisition module 100 is used to acquire two-dimensional echo data from the corner reflector; The initial correction module 200 is used to calculate the amplitude and phase errors of each radar channel relative to the reference channel based on the two-dimensional echo data, and generate a compensation matrix; and to correct the original echo signal of the two-dimensional echo data according to the compensation matrix to obtain the initial correction data. The distance filtering module 300 is used to remove invalid distance cells from the initial correction data based on a preset energy threshold, select and sort the remaining cells according to the calculated signal-to-noise ratio and normalized amplitude variance, eliminate the linear phase shift of the distance cells and align the strongest scattering point to the center, iteratively rotate the initial phase of each distance cell to make their directions consistent, and obtain the first processed data. The iteration module 400 is used to divide the distance units in the first processed data according to a preset variance threshold to obtain high signal-to-noise ratio (SNR) distance units and low SNR distance units; the distance unit ranked first in the first processed data is used as the initial distance unit for iteration; during iteration, the fixed initial phase is expanded to obtain the expanded phase, and then the high SNR distance units and low SNR distance units are distinguished and calculated to obtain the corresponding phase noise variances respectively; based on the expanded phase and the phase noise variance obtained in each iteration, the channel phase error is estimated by weighted least squares method. The output module 500 is used to obtain output data after all distance units in the first processed data have been iterated.

[0066] The functions of each module in the above-mentioned ground-based MIMO SAR amplitude and phase error correction device correspond to the steps in the above-mentioned ground-based MIMO SAR amplitude and phase error correction method embodiment, and their functions and implementation processes will not be described in detail here.

[0067] This application also provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the ground-based MIMO SAR amplitude and phase error correction method as described in any of the above embodiments.

[0068] This application also provides a computer-readable storage medium storing a program. The computer-readable storage medium refers to a data storage medium, which may include, but is not limited to, floppy disks, optical disks, hard disks, flash memory, USB flash drives, and / or Memory Sticks. The computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The working process, details, and technical effects of the computer-readable storage medium provided in this embodiment can be found in the above embodiment regarding a method for correcting amplitude and phase errors of ground-based MIMO SAR, and will not be repeated here.

[0069] The application also provides a computer program product, including computer instructions that, when executed by a processor, implement the steps of the ground-based MIMO SAR amplitude and phase error correction method as described in any of the above embodiments.

[0070] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), RAMbus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and RAMbus dynamic RAM (RDRAM).

[0071] The technical features of the above embodiments can be combined arbitrarily. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as the combination of these technical features does not contradict each other, it should be considered within the scope of this specification. The above embodiments only illustrate several implementation methods of this application, and their descriptions are relatively specific and detailed, but they should not be construed as limiting the scope of the invention patent. It should be noted that for those skilled in the art, several modifications and improvements can be made without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this patent application should be determined by the appended claims.

Claims

1. A method for correcting amplitude and phase errors in ground-based MIMO SAR, characterized in that, include: Two-dimensional echo data is acquired from a corner reflector; The amplitude and phase errors of each radar channel relative to the reference channel are calculated based on the two-dimensional echo data, and a compensation matrix is ​​generated. The original echo signal of the two-dimensional echo data is corrected according to the compensation matrix to obtain the initial corrected data; Based on the preset energy threshold, invalid range cells in the initial correction data are removed. The remaining cells are selected and sorted according to the calculated signal-to-noise ratio and normalized amplitude variance. The linear phase shift of the range cells is eliminated and the strongest scattering point is aligned to the center. The initial phase of each range cell is iteratively rotated to make their directions consistent, and the first processed data is obtained. The distance units in the first processed data are divided according to a preset variance threshold to obtain high signal-to-noise ratio distance units and low signal-to-noise ratio distance units; The distance unit ranked first in the first processed data is used as the initial distance unit, and the process is iterated; During iteration, the phase of the fixed initial phase is expanded to obtain the expanded phase. Then, the high signal-to-noise ratio range cells and the low signal-to-noise ratio range cells are distinguished and calculated to obtain the corresponding phase noise variance. Based on the expanded phase and the phase noise variance obtained in each iteration, the channel phase error is estimated by weighted least squares method. The channel phase error is used to compensate for the error in the first processed data in each iteration; Once all distance units in the first processed data have been iterated, the output data is obtained.

2. The method for correcting amplitude and phase errors of ground-based MIMO SAR according to claim 1, characterized in that, The preset variance threshold is 0.

27.

3. The method for correcting amplitude and phase errors of ground-based MIMO SAR according to claim 1, characterized in that, The step of dividing the distance units in the first processed data according to a preset variance threshold to obtain high signal-to-noise ratio distance units and low signal-to-noise ratio distance units includes: The distance cells in the first processed data whose normalized amplitude variance is less than a preset variance threshold are used as high signal-to-noise ratio distance cells. The distance cells in the first processed data whose normalized amplitude variance is greater than a preset variance threshold are used as low signal-to-noise ratio distance cells.

4. The method for correcting amplitude and phase errors of ground-based MIMO SAR according to claim 3, characterized in that, The step of distinguishing and calculating the high signal-to-noise ratio range cells and low signal-to-noise ratio range cells to obtain the corresponding phase noise variance includes: For distance cells in the first processed data whose normalized amplitude variance is less than a preset variance threshold, the phase noise variance is calculated using a signal-to-noise ratio analytical model. The analytical model calculates the phase noise variance through the quantitative relationship of the signal-to-noise ratio. During iteration, for distance cells in the first processed data whose normalized amplitude variance is greater than a preset variance threshold, the phase noise variance is calculated using an iterative model based on residual error. The iterative model updates the current phase noise variance using the residual error estimate from the previous iteration step.

5. The method for correcting amplitude and phase errors of ground-based MIMO SAR according to claim 1, characterized in that, The estimation of channel phase error using weighted least squares based on the expanded phase and phase noise variance obtained in each iteration includes: A weight matrix is ​​constructed using the reciprocal of the phase noise variance as the weighting coefficient; Based on the weight matrix, the expanded phase is weighted and averaged to obtain the residual phase error.

6. The method for correcting amplitude and phase errors of ground-based MIMO SAR according to claim 5, characterized in that, The step of using the channel phase error to compensate for errors in the first processed data of each iteration includes: The residual phase error is used to compensate for errors in the first processed data of each iteration.

7. A ground-based MIMO SAR amplitude and phase error correction device, characterized in that, include: The data acquisition module is used to acquire two-dimensional echo data from the corner reflector; The initial correction module is used to calculate the amplitude and phase errors of each radar channel relative to the reference channel based on the two-dimensional echo data, and generate a compensation matrix. The original echo signal of the two-dimensional echo data is corrected according to the compensation matrix to obtain the initial corrected data; The distance filtering module is used to remove invalid distance cells from the initial correction data based on a preset energy threshold, select and sort the remaining cells according to the calculated signal-to-noise ratio and normalized amplitude variance, eliminate the linear phase shift of the distance cells and align the strongest scattering point to the center, iteratively rotate the initial phase of each distance cell to make their directions consistent, and obtain the first processed data. The iteration module is used to divide the distance units in the first processed data according to a preset variance threshold to obtain high signal-to-noise ratio distance units and low signal-to-noise ratio distance units; The distance unit ranked first in the first processed data is used as the initial distance unit for iteration. During iteration, the phase of the fixed initial phase is expanded to obtain the expanded phase. Then, the high signal-to-noise ratio distance unit and the low signal-to-noise ratio distance unit are distinguished and calculated to obtain the corresponding phase noise variance. Based on the expanded phase and the phase noise variance obtained in each iteration, the channel phase error is estimated by weighted least squares method. The output module is used to obtain output data after all distance units in the first processed data have been iterated.

8. A computer device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the method for correcting the amplitude and phase error of ground-based MIMO SAR as described in any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a program, wherein when the program is executed by a processor, it implements a method for correcting the amplitude and phase errors of ground-based MIMO SAR as described in any one of claims 1 to 6.

10. A computer program product comprising computer instructions, characterized in that, When executed by a processor, the computer instructions implement the steps of the method for correcting amplitude and phase errors of ground-based MIMO SAR as described in claims 1 to 6.

Images