Microwave correlated imaging method and device based on frequency and phase joint modulation

Abstract

The invention discloses a microwave correlated imaging method and device based on frequency and phase joint modulation, and relates to the technical field of radars. The method comprises the following steps: acquiring radar parameters; performing grid division on the target imaging scene; constructing a random radiation field of frequency and phase joint modulation based on radar parameters, a grid division result and a frequency and phase modulation function; receiving an echo signal of the target by using the random radiation field, and obtaining a distance pulse pressure envelope based on the received echo signal; and on the basis of the sub-optimization problems of the random radiation field, the distance pulse pressure envelope and the noise variance, the sub-optimization problem of the target sparse characteristic parameter and the sub-optimization problem of the super-resolution image reconstruction, the super-resolution image of the target is determined by alternately estimating the noise variance, the target sparse characteristic parameter and the super-resolution image. The method can effectively solve the problems that a random radiation field is single in construction mode and insufficient in flexibility, the image super-resolution capability is remarkably reduced under the condition of a low signal-to-noise ratio, and the imaging quality is limited.

Description

Microwave Correlation Imaging Method and Device Based on Frequency-Phase Joint Modulation Technical Field

[0001] This invention belongs to the field of radar technology, specifically relating to a microwave correlation imaging method and device based on frequency-phase joint modulation. Background Technology

[0002] Microwave correlation imaging is an imaging method based on wavefront modulation and spatiotemporal random radiation fields. Its basic idea is to randomly modulate the transmitted signal, generating independent random radiation fields in both spatial and temporal dimensions. Multiple observations of the target scene are then performed, and the received echoes are correlated with the corresponding random radiation fields to extract the target's scattering information, achieving high-resolution imaging. Compared to traditional radar imaging techniques, this method does not rely on array physical aperture expansion, nor does it require Doppler information generated by the relative motion between the platform and the target. It exhibits better anti-interference capabilities and higher imaging resolution in complex electromagnetic environments and has shown application potential in fields such as precision guidance, Earth staring imaging, distributed satellite observation, and security inspection.

[0003] Despite the progress made in microwave correlation imaging technology in recent years, its engineering applications still face several key challenges. First, microwave correlation imaging is highly dependent on the stochastic characteristics of the spatiotemporally incoherent radiation field. However, the construction methods of random radiation fields in practical systems are usually limited and constrained by hardware conditions such as signal dimension and array element size, making it difficult to achieve ideal randomness, thus weakening imaging performance. Second, because this type of method relies on spatiotemporally incoherent random radiation fields to decouple target information within the beam, the echo signals between different pulses differ significantly, making effective coherent accumulation difficult. This results in low echo energy utilization and limited signal-to-noise ratio gain, thereby affecting imaging quality.

[0004] To address the aforementioned issues, existing research primarily improves microwave correlation imaging performance in two ways: firstly, by enhancing the randomness of the random radiation field generation method; and secondly, by combining advanced signal processing and imaging reconstruction algorithms to improve imaging quality. Regarding random radiation field design, some methods rely on complex hardware structures, resulting in high system implementation costs. Furthermore, existing research often employs a single modulation scheme (such as random frequency modulation), limiting its adaptability to complex scenarios. In terms of imaging reconstruction, while introducing correlation processing techniques can improve resolution to some extent, the potential for performance improvement is limited. Compressed sensing-based methods are highly sensitive to model parameters and prior assumptions; improper parameter selection can easily lead to image degradation. Deep learning methods, on the other hand, rely on large amounts of training data, and the data distribution varies significantly across different scenarios, leaving room for improvement in model generalization ability and cross-scenario applicability. For example, the paper titled "A Sparse Target Microwave Correlation Imaging Method Based on an Improved Orthogonal Matching Pursuit (OMP) Algorithm" proposes an improved OMP algorithm for sparse target microwave correlation imaging, incorporating the concept of frequency agility. First, combining the idea of ​​frequency agility, the frequency of each transmitted signal randomly agilely varies between pulses, forming a random radiation field with two-dimensional incoherent characteristics in space and time. Then, the least squares solution step in the OMP algorithm is improved using the conjugate gradient method, reducing the computational load. Finally, comparative experiments verify the algorithm's performance. However, this method has two shortcomings. First, the construction method of the random radiation field is too simplistic and lacks flexibility. Second, the influence of noise on the model is not considered during the modeling and solution process, which cannot guarantee the performance of microwave correlation imaging at low signal-to-noise ratios. A paper titled "A Microwave Correlation Forward-Looking Imaging Method Based on TSVDT" proposes a new microwave correlation forward-looking imaging method optimized for the pseudo-inverse algorithm. This method first acquires a two-dimensional spatiotemporal random radiation field based on a frequency-hopping radar array. Then, it combines truncated singular value decomposition (TSVD) with Tikhonov regularization to propose a joint TSVD-Tikhonov (TSVDT) processing method to improve the spatiotemporal random radiation array matrix, ultimately achieving microwave correlated imaging. Although this method improves imaging stability and anti-interference capability to some extent, it is essentially still a pseudo-inverse solution framework, and its imaging performance is highly dependent on the singular value distribution of the random radiation array matrix. When the randomness of the radiation field is insufficient, the array size is limited, or the model is mismatched, matrix ill-conditioning is still difficult to avoid. Furthermore, under low signal-to-noise ratio conditions, super-resolution imaging performance still faces certain bottlenecks, affecting subsequent radar signal processing performance.

[0005] In other words, existing related technologies suffer from problems such as a single method of constructing random radiation fields, insufficient flexibility, and a significant decrease in image super-resolution capability and limited imaging quality under low signal-to-noise ratio conditions. Summary of the Invention

[0006] To address the aforementioned problems in the prior art, this invention provides a microwave correlation imaging method and device based on frequency-phase joint modulation. The technical problem to be solved by this invention is achieved through the following technical solution: This invention provides a microwave correlation imaging method based on frequency-phase joint modulation, comprising: acquiring radar parameters; dividing the target imaging scene into grids to obtain multiple spatial angles; constructing a frequency-phase joint modulated random radiation field based on the radar parameters, the multiple spatial angles, a frequency modulation function, and a phase modulation function; receiving the target's echo signal using the frequency-phase joint modulated random radiation field, and performing range compression processing on the received echo signal to obtain a range pulse compression envelope; based on the frequency-phase joint modulated random radiation field, the range pulse compression envelope, a sub-optimization problem of noise variance, a sub-optimization problem of target sparse characteristic parameters, and a sub-optimization problem of super-resolution image reconstruction combining the noise variance and the target sparse characteristic parameters, by alternately estimating the noise variance, the target sparse characteristic parameters, and the super-resolution image, determining a super-resolution image of the target that meets preset requirements.

[0007] The present invention also provides a microwave correlation imaging device based on frequency-phase joint modulation, including a processor, a communication interface, a memory, and a communication bus. The processor, the communication interface, and the memory communicate with each other through the communication bus. The memory is used to store computer programs. When the processor executes the program stored in the memory, it implements the steps of the microwave correlation imaging method based on frequency-phase joint modulation described above.

[0008] Compared with the prior art, the beneficial effects of the present invention are as follows: 1) High flexibility in constructing random radiation fields. The present invention constructs a random radiation field suitable for both ideal full arrays and sparse arrays through frequency-phase joint modulation, which significantly improves the flexibility and randomness of the random radiation field in the spatial and temporal domains, and overcomes the problems of single modulation form and insufficient adaptability in traditional microwave correlation imaging.

[0009] 2) Improved robustness of the imaging algorithm. This invention considers the impact of noise on microwave correlation imaging performance, and integrates super-resolution images, noise statistical parameters (i.e., noise variance), and target sparsity parameters into the same imaging model. It also uses an alternating optimization method to adaptively estimate these parameters, avoiding dependence on fixed noise models or prior parameters, thereby improving the robustness of the imaging algorithm.

[0010] 3) High-performance super-resolution imaging for low signal-to-noise ratio scenarios is achieved. This invention effectively suppresses noise interference during the optimization process by modeling and iteratively approximating the statistical characteristics of noise. It can still achieve high-performance super-resolution imaging under low signal-to-noise ratio conditions, effectively improving the efficiency of echo information utilization and imaging quality, and enhancing the applicability of microwave correlation imaging in complex environments.

[0011] The present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. Attached Figure Description

[0012] Figure 1 is a flowchart illustrating the microwave correlation imaging method based on frequency-phase joint modulation provided in an embodiment of the present invention; Figure 2 is a schematic diagram illustrating the arrangement of array elements and coordinate system relationship in an exemplary phased array provided in an embodiment of the present invention; Figure 3 is a schematic diagram illustrating the specific arrangement of an exemplary phased array provided in an embodiment of the present invention; Figure 4A is a two-dimensional radiation pattern of an exemplary frequency-phase joint modulated random radiation field at random time 1 within the beam illumination time provided in an embodiment of the present invention; Figure 4B is a three-dimensional radiation pattern of an exemplary frequency-phase joint modulated random radiation field at random time 1 within the beam illumination time provided in an embodiment of the present invention; Figure 5A is an exemplary frequency-phase joint... Figure 5B is a two-dimensional radiation pattern of the modulated random radiation field at a random time 2 within the beam illumination time, provided by an embodiment of the present invention; Figure 6A is an exemplary target scene diagram provided by an embodiment of the present invention when the angular interval between different targets is 4°; Figure 6B is an exemplary schematic diagram of the real aperture scanning imaging result in the scene shown in Figure 6A, provided by an embodiment of the present invention; Figure 6C is an exemplary schematic diagram of the imaging result of the SPGL1 algorithm in the scene shown in Figure 6A, provided by an embodiment of the present invention; Figure 6D is an exemplary schematic diagram of the imaging result of the SPGL1 algorithm in the scene shown in Figure 6A, provided by an embodiment of the present invention. Below are schematic diagrams of the imaging results of the method proposed in this invention; Figure 7A is an exemplary target scene diagram provided by an embodiment of this invention when the angular interval between different targets is 2°; Figure 7B is an exemplary schematic diagram of the real aperture scanning imaging results provided by an embodiment of this invention in the scene shown in Figure 7A; Figure 7C is an exemplary schematic diagram of the imaging results of the SPGL1 algorithm provided by an embodiment of this invention in the scene shown in Figure 7A; Figure 7D is an exemplary schematic diagram of the imaging results of the method proposed in this invention in the scene shown in Figure 7A; Figure 8A is an exemplary target scene diagram provided by an embodiment of this invention when the angular interval between different targets is 1°; Figure 8B ...C is an exemplary schematic diagram of the imaging results of the SPGL1 algorithm provided by an embodiment of this invention; Figure 7D is an exemplary schematic diagram of the imaging results of the method proposed in this invention in the scene shown in Figure 7A; Figure 8A is an exemplary target scene diagram provided by an embodiment of this invention when the angular interval between different targets is 1°; Figure 8B is an exemplary schematic diagram of the imaging results of the method proposed in this invention in the scene shown in Figure 7A; Figure 8C is an exemplary schematic diagram of the imaging results of the SPGL1 algorithm provided by an embodiment of this invention; Figure 8D is an exemplary schematic diagram of the imaging results of the method proposed in this invention in the scene shown in Figure 7A; Figure 8B is an exemplary schematic diagram of the The embodiments provide an exemplary schematic diagram of the real aperture scanning imaging result in the scenario shown in Figure 8A; Figure 8C is an exemplary schematic diagram of the imaging result of the SPGL1 algorithm in the scenario shown in Figure 8A; Figure 8D is an exemplary schematic diagram of the imaging result of the method proposed in this invention in the scenario shown in Figure 8A; Figure 9A is an exemplary schematic diagram of the imaging result of the SPGL1 algorithm when the angular interval between different targets is 2° and the SNR is 5dB; Figure 9B is an exemplary schematic diagram of the imaging result of the method proposed in this invention when the angular interval between different targets is 2° and the SNR is 5dB.Figure 9C is an exemplary imaging result diagram of the SPGL1 algorithm when the angular interval between different targets is 2° and the SNR is -5dB, provided by an embodiment of the present invention; Figure 9D is an exemplary imaging result diagram of the method proposed in this invention when the angular interval between different targets is 2° and the SNR is -5dB, provided by an embodiment of the present invention; Figure 9E is an exemplary imaging result diagram of the SPGL1 algorithm when the angular interval between different targets is 2° and the SNR is -15dB, provided by an embodiment of the present invention; Figure 9F is an exemplary imaging result diagram of the method proposed in this invention when the angular interval between different targets is 2° and the SNR is -15dB, provided by an embodiment of the present invention. Detailed Implementation

[0013] The present invention will be further described in detail below with reference to specific embodiments, but the implementation of the present invention is not limited thereto.

[0014] Figure 1 is a flowchart of a microwave correlation imaging method based on frequency-phase joint modulation provided by an embodiment of the present invention. As shown in Figure 1, the method includes: S101, acquiring radar parameters.

[0015] Radar parameters include: radar electromagnetic parameters, phased array structural parameters, and radar pulse repetition count. Among them, radar electromagnetic parameters include: carrier frequency ,wavelength Pulse width and signal bandwidth The structural parameters of a phased array include: the total number of elements in the phased array. And the arrangement of each array element. For example, when the antenna array surfaces of a phased array are distributed in the O-xy plane of a Cartesian coordinate system, the first... The arrangement of the array elements can be represented as follows: ,and and The first The x and y coordinates of each array element in this coordinate system. It is a positive integer, and, A phased array can be either an ideal full array or a sparse array. An ideal full array refers to an array where the element spacing is 1:1. A phased array.

[0016] S102. Divide the target imaging scene into a grid to obtain multiple spatial angles.

[0017] For example, the target imaging scene is divided into multiple grids using preset azimuth and pitch angle intervals. Each grid represents a spatial angle, and each spatial angle corresponds to an azimuth and a pitch angle. For instance, when the scene is divided into grids... At the first spatial angle, the first... The azimuth and elevation angles corresponding to each spatial angle are respectively expressed as follows: and , .

[0018] S103. Based on radar parameters, multiple spatial angles, frequency modulation function, and phase modulation function, a random radiation field with joint frequency and phase modulation is constructed.

[0019] In this invention, at least one of the frequency modulation function and the phase modulation function is an independent random process function that varies with time. This independent random process function refers to a function with time as the independent variable, where each time point has a corresponding random value, and the changes in random values ​​across different and non-overlapping time intervals are independent of each other. For example, both the frequency modulation function and the phase modulation function can be functions whose function values ​​are independent and both follow a uniform distribution.

[0020] S104. The target echo signal is received using a random radiation field modulated by frequency and phase, and the received echo signal is subjected to range compression processing to obtain the range pulse compression envelope.

[0021] It should be noted that the random radiation field of frequency-phase joint modulation remains unchanged during each pulse transmission and reception, but varies between different pulses. In this way, multiple independent random samples can be provided for the radar system.

[0022] S105. Sub-optimization problems based on frequency-phase joint modulation of random radiation field, range pulse compression envelope, noise variance, target sparse characteristic parameters, and super-resolution image reconstruction based on joint noise variance and target sparse characteristic parameters. By alternately estimating noise variance, target sparse characteristic parameters and super-resolution image, a super-resolution image of the target that meets the preset requirements is determined.

[0023] In some embodiments, S103 can be specifically implemented as follows: based on radar electromagnetic parameters, the arrangement position of each array element, the number of radar pulse repetitions, multiple spatial angles, antenna beam pointing, frequency modulation function, and phase modulation function, a random radiation field matrix jointly modulated by the frequency and phase of different pulses is constructed, wherein the first dimension of the random radiation field matrix jointly modulated by the frequency and phase of different pulses is the number of radar pulse repetitions. The second dimension is the total number of spatial angles. Furthermore, the random radiation field matrix jointly modulated by the frequency and phase of different pulses represents the random radiation field jointly modulated by the frequency and phase.

[0024] For example, the constructed frequency-phase jointly modulated random radiation field This can be expressed as the following formulas (1) to (3): (1); (2); (3); where, according to formulas (1) and (2), it can be seen that, yes OK A matrix of columns, express The row element, It is a positive integer. , express Any element in, and They represent the first The azimuth and elevation angles corresponding to each spatial angle For example, when and When all values ​​are 1, That is The same applies to the rest. This represents the total number of elements in a phased array. This refers to the pointing position of the antenna beam, and and These represent the azimuth and elevation angles, respectively. Indicates the first pulse At the time The amplitude of the excitation current on each array element and They represent At the time The modulation frequency and modulation phase added to each array element Represents the speed of light. The imaginary unit, Represents the sine function. Represents the cosine function. This represents the natural exponential function. The value can be set according to actual needs. For example, it can be set to 1. This invention does not impose specific limitations on its value.

[0025] In some embodiments, the construction method of the sub-optimization problem of noise variance, the sub-optimization problem of target sparse characteristic parameters and the sub-optimization problem of super-resolution image reconstruction in S105 above is as follows: S1, construct a microwave correlation imaging model of the random radiation field of frequency-phase joint modulation based on the random radiation field of frequency-phase joint modulation, the range pulse compression envelope and the phase matrix, wherein the microwave correlation imaging model includes a noise vector and a super-resolution image to be reconstructed.

[0026] For example, the expressions for the microwave correlation imaging model are as follows: (4)~(5) (4); (5); among them, Indicates distance pulse pressure envelope The vector corresponding to the distance profile of the imaging plane. This represents a matrix vectorization operator, used to convert a matrix into a vector. express hour The corresponding value in the middle. Let be the phase matrix, and , This represents the slant range difference between the slant profile corresponding to the imaging plane and the reference slant range. This represents matrix operation functions used to create diagonal matrices or extract diagonal elements from a matrix. This represents the super-resolution image that needs to be reconstructed. Represents the noise vector; The transpose symbol indicates the transpose of a matrix.

[0027] S2. Based on the Maximum A Posteriori (MAP) estimation and maximum likelihood estimation criteria, the microwave correlation imaging model is transformed into a multi-parameter optimization problem under sparse constraints. The multi-parameter optimization problem includes the super-resolution image to be reconstructed, the noise variance to be solved, and the target sparse characteristic parameters to be solved.

[0028] Since the noise follows an independent complex Gaussian distribution, the image exhibits significant sparsity, and the two-dimensional image follows a Laplace distribution. Therefore, based on the MAP estimation and maximum likelihood estimation criteria, the high-resolution image reconstruction problem of a random radiation field with joint frequency and phase modulation can be transformed into a multi-parameter optimization problem under sparse constraints. For example, the expression for the multi-parameter optimization problem under sparse constraints is Equation (6): (6); among them, This represents the super-resolution image that needs to be solved. This represents the noise variance that needs to be solved. This represents the target sparse property parameter that needs to be solved. The super-resolution image representing the optimization variable when the optimization problem reaches its minimum value. Noise variance and target sparsity parameters The value of , Represents the natural logarithm function. express The absolute value, and Represents super-resolution images The Middle Units, Represents the target sparsity parameter The first included There are Laplace scaling parameters, where... yes The corresponding Laplacian scaling parameters, Represents the L2 norm. This represents the square of the L2 norm.

[0029] S3. Decompose the multi-parameter optimization problem into a sub-optimization problem of noise variance, a sub-optimization problem of target sparse characteristic parameters, and a sub-optimization problem of super-resolution image reconstruction.

[0030] Due to the super-resolution image in the above formula (6) Target sparsity characteristics parameters Noise variance Since all parameters are unknown, the multi-parameter optimization problem can be decomposed into the following three sub-optimization problems: the sub-optimization problem of noise variance, as shown in formula (7): (7); Target sparsity parameter The sub-optimization problem is as shown in formula (8): (8); The sub-optimization problem of super-resolution image reconstruction is as shown in formula (9): (9).

[0031] Here, by using the optimization method to solve the sub-optimization problem of the noise variance shown in the above formula (7), the expression for the noise variance in the scene can be obtained as formula (10): (10).

[0032] Similarly, by using optimization methods to solve the sub-optimization problem of the target sparse characteristic parameters shown in the above formula (8), the expression for the target sparse characteristic parameters in the scene can be obtained as formula (11): (11), where, express The first included Laplace scale parameter It is a small constant. The specific value can be set according to actual needs, and is usually taken as... The average is 5%-10%, therefore, when As the parameter estimates change with each iteration, Also with each iteration, the estimated Make changes.

[0033] In some embodiments, the above-mentioned S105 is implemented through steps S1051 to S1053: S1051, in the process of calculating noise variance, target sparsity parameter and super-resolution image... When performing the alternating estimation, obtain the first... The super-resolution image estimated in the first step, based on the first... The super-resolution image estimated in the first estimation, the sub-optimization problem of noise variance, the random radiation field of frequency-phase joint modulation, and the range pulse compression envelope are used to determine the second... The noise variance estimated in the first estimation, and based on the first... The sub-optimization problem of estimating the super-resolution image and the sparse characteristic parameters of the target determines the first... The target sparse property parameters are estimated in the second step; It is a positive integer greater than or equal to 1, when When =1, the first The super-resolution image estimated in the first estimation is the super-resolution image estimated in the 0th estimation, which is the initial super-resolution image generated in advance.

[0034] In some embodiments, when =1, meaning that before the first alternating estimation of noise variance, target sparsity parameters, and the super-resolution image, the microwave correlation imaging model described above can be solved using the least squares method to obtain a super-resolution image. This obtained super-resolution image is then used as the initial super-resolution image for the first alternating estimation of noise variance, target sparsity parameters, and the super-resolution image. In other embodiments, other methods can also be used to pre-generate the initial super-resolution image. When If the value is greater than 1, then when performing the second or more alternating estimations of noise variance, target sparsity parameters and super-resolution image, the super-resolution image estimated in the previous time is directly obtained and used for the alternating estimation of noise variance, target sparsity parameters and super-resolution image in this current iteration.

[0035] In the process of analyzing noise variance, target sparsity parameters, and super-resolution images... When performing the alternating estimation, the first step is to determine the... The noise variance of the first estimate and the second estimate The target sparsity property parameters are estimated in the second estimation. It should be noted that determining the third... The noise variance of the first estimate and the second estimate The steps for estimating the target sparse property parameters can be performed sequentially or simultaneously; this invention does not limit this.

[0036] Specifically, in determining the first When estimating the noise variance for the second time, the random radiation field jointly modulated by frequency and phase, the range pulse compression envelope, and the first time are considered. Substituting the estimated super-resolution image into formula (10) above, we can obtain formula (12): (12), where, Indicates the first The super-resolution image estimated in the first step can be calculated using formula (12). The noise variance of the second estimate Similarly, in determining the first... When estimating the target sparse characteristic parameters for the second time, the random radiation field jointly modulated by frequency and phase, the range pulse compression envelope, and the first time... Substituting the estimated super-resolution image into formula (11) above, we can obtain formula (13): (13), where, express The Middle Units, Indicates according to The obtained number Second The first number can be calculated using formula (13). The target sparsity parameter is estimated in the second step. .

[0037] S1052, Based on the random radiation field jointly modulated by frequency and phase, the range pulse compression envelope, and the first... The noise variance of the first estimate, the second estimate The target sparse characteristic parameters are estimated for the first time, and the sub-optimization problem of super-resolution image reconstruction is used to determine the first... Super-resolution image estimated at the second time.

[0038] Specifically, the random radiation field, range pulse compression envelope, and the first frequency-phase co-modulated random radiation field are combined. The noise variance of the first estimate, the second estimate The estimated target sparse characteristic parameters are substituted into the sub-optimization problem of super-resolution image reconstruction. The derivative of this sub-optimization problem is then obtained to yield the equation to be solved. An optimization algorithm (e.g., the conjugate gradient method) is used to iteratively solve the equation to obtain the ... Super-resolution image estimated at the second time.

[0039] For example, the random radiation field, range pulse compression envelope, and the first frequency-phase co-modulated random radiation field are used. The noise variance of the first estimate, the second estimate After substituting the estimated target sparsity parameter into the above formula (9), we can obtain formula (14): (14), then, differentiate formula (14) and then... Substituting these equations into the equation, we obtain the equation to be solved as shown in formula (15): (15), where, express The first in Laplace scale parameter Let represent a diagonal matrix, and The The elements are , It is a diagonal matrix, and The The elements are , It is a non-negative, tiny component, in order to avoid It is not differentiable at 0, and its specific value can be set according to actual needs. This is the conjugate transpose symbol, indicating the conjugate transpose. By using the conjugate gradient method to solve the equation shown in formula (15), the first... Super-resolution image estimated at the second time It should be noted that other optimization algorithms can also be used to solve the equation shown in formula (15), and this invention does not limit this.

[0040] S1053, Determine the first If the estimated super-resolution image meets the preset requirements, then the first estimated super-resolution image will be... The estimated super-resolution image is taken as the super-resolution image of the target that meets the preset requirements; if it does not meet the requirements, then let After that, we continue to analyze the noise variance, target sparsity parameters, and the super-resolution image. The estimation is repeated several times until a super-resolution image of the target that meets the preset requirements is obtained.

[0041] Specifically, determine the first Super-resolution image estimated at the second time With the Super-resolution image estimated at the second time The gap between them; based on the gap and the preset threshold The size relationship between them determines the first Super-resolution image estimated at the second time Does it meet the preset requirements, where the difference is less than or equal to the preset threshold? When, it indicates the first Super-resolution image estimated at the second time The preset requirements are met, but when the difference exceeds the preset threshold... When, it indicates the first Super-resolution image estimated at the second time The preset requirements are not met. For example, the first... Super-resolution image estimated at the second time With the Super-resolution image estimated at the second time The difference between them can be expressed as , and, when When, it indicates the first Super-resolution image estimated at the second time If the preset requirements are met, otherwise, it indicates that the first... Super-resolution image estimated at the second time The preset requirements are not met. When the first... Super-resolution image estimated at the second time If the preset requirements are not met, then... Then, the process returns to step S1051 above to continue alternating estimation of noise variance, target sparsity parameters, and super-resolution image until a super-resolution image meeting preset requirements is obtained. It should be noted that the preset threshold... The threshold discrimination constant is relatively small, and its specific value can be set according to actual needs. This invention does not limit it.

[0042] It should be understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0043] As can be seen from the above, addressing the limitations of traditional microwave correlation imaging methods in terms of their singular random radiation field construction and performance constraints under low signal-to-noise ratio conditions, this invention significantly enhances the randomness and flexibility of the random radiation field by constructing a frequency-phase joint modulation random radiation field model suitable for both ideal full-array and sparse arrays. Simultaneously, this invention establishes a multi-parameter adaptive alternating estimation imaging framework, unifying super-resolution images, noise statistical parameters, and target sparse characteristic parameters into a single imaging model. Through alternating optimization, multiple parameters are adaptively estimated, enabling accurate estimation of target sparse characteristic parameters and noise feature parameters during the alternating estimation process. This better distinguishes the influence of signal and noise, improves target imaging resolution, and reduces noise interference during optimization. This method significantly enhances the robustness and applicability of microwave correlation imaging under complex environments and low signal-to-noise ratio conditions, possessing high engineering application value.

[0044] The present invention also provides a microwave correlation imaging device based on frequency-phase joint modulation, including a processor, a communication interface, a memory, and a communication bus. The processor, the communication interface, and the memory communicate with each other through the communication bus. The memory is used to store computer programs. When the processor executes the program stored in the memory, it implements the steps of the microwave correlation imaging method based on frequency-phase joint modulation described above.

[0045] It should be noted that the above-mentioned device can be any processing device capable of implementing the above-mentioned method, and the specific type of the device is not limited in this embodiment of the invention. The processor can be a Central Processing Unit (CPU), but it can also be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor can be a microprocessor or any conventional processor. In some embodiments, the memory can be an internal storage unit, such as a hard disk or RAM; in other embodiments, it can be an external storage device, such as a plug-in hard disk, a smart memory card (SMC), a secure digital card (SD), a flash card, etc. Furthermore, the memory can include both internal storage units and external storage devices. The memory is used to store the operating system, applications, boot loader, data, and other programs, such as the program code of the computer program. The memory can also be used to temporarily store data that has been output or will be output.

[0046] This invention also provides a computer-readable storage medium storing a computer program that, when executed by a processor, can implement the steps described in the various method embodiments above.

[0047] This invention also provides a computer program product that, when run on an electronic device, enables the electronic device to perform the steps described in the various method embodiments above.

[0048] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the methods of the above embodiments of the present invention can be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium can include at least: any entity or device capable of carrying computer program code to an electronic device, a recording medium, a computer memory, a read-only memory (ROM), a random access memory (RAM), an electrical carrier signal, a telecommunication signal, and a software distribution medium. Examples include USB flash drives, portable hard drives, magnetic disks, or optical disks.

[0049] In the above embodiments, the descriptions of each embodiment have different focuses. For parts that are not described in detail or recorded in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0050] The following simulation experiments verify the beneficial effects of the present invention: 1) Radar electromagnetic parameters and phased array structural parameters: In this simulation experiment, the center frequency is used as the reference. 17GHz, wavelength for With a bandwidth of 2GHz and a pulse width of for The distance from the center of the scene is Ideal distance resolution is Taking a radar with a spatial angular resolution of 3.2° as an example, the basic parameters of the radar are shown in Table 1. The arrangement of the array elements in the phased array and their relationship to the coordinate system are shown in Figure 2. In this figure, each black circle represents an array element, and each solid black dot represents a far-field observation point, which corresponds to an azimuth angle. and a pitch angle The ideal element spacing is The aperture size of the phased array is Array sparsity The value is 0.5, representing the number of phased array elements. The phased array has 648 elements. The specific structural parameters of the phased array are shown in Table 2, and the specific arrangement of the phased array is shown in Figure 3. Each black circle represents an array element.

[0051] Table 1

[0052] Table 2

[0053] 2) Constructing a random radiation field with joint frequency and phase modulation: (2a) Setting the first Modulation frequency added to each element and modulation phase They are mutually independent and all follow a uniform distribution, and, , ,in, Indicates the interval as Uniform distribution Indicates the interval as Uniform distribution express obey The uniform distribution represented Similarly.

[0054] (2b) Construct a frequency-phase co-modulated random radiation field in conjunction with (2a). For example, Figures 4A, 4B, 5A, and 5B are two-dimensional and three-dimensional radiation patterns of the frequency-phase co-modulated random radiation field at two randomly selected moments within the beam illumination time. In Figures 4A and 5A, the vertical axis represents the elevation angle, and the horizontal axis represents the azimuth angle. In Figures 4B and 5B, the x-coordinate represents the elevation angle, the y-coordinate represents the azimuth angle, and the z-coordinate represents the normalized amplitude. The elevation angle can be expressed as... The unit is degrees, and the azimuth angle can be expressed as... The unit is degrees. Specifically, Figure 4A is a two-dimensional radiation pattern of the frequency-phase co-modulated random radiation field at a random moment (e.g., referred to as random moment 1) within the beam illumination time; Figure 4B is a three-dimensional radiation pattern of the frequency-phase co-modulated random radiation field at random moment 1 within the beam illumination time; Figure 5A is a two-dimensional radiation pattern of the frequency-phase co-modulated random radiation field at another random moment (e.g., referred to as random moment 2) within the beam illumination time; and Figure 5B is a three-dimensional radiation pattern of the frequency-phase co-modulated random radiation field at random moment 2 within the beam illumination time.

[0055] 3) Based on the microwave correlation imaging model of the random radiation field with joint frequency and phase modulation and the multi-parameter optimization problem under sparse constraints, as well as the alternating estimation and solution of multiple sub-optimization problems, the final super-resolution image is obtained.

[0056] 4) Simulation Result Analysis: The effect of the super-resolution image reconstructed using the method proposed in this invention is measured by the Mean Squared Error (MSE). MSE is expressed as: In the formula, and The first two images represent the amplitude-normalized reference image and the super-resolution image reconstructed using the method proposed in this invention, respectively. The pixel (i.e., the first pixel) (Units). Under the same observation scenario, the smaller the MSE, the closer the calculated target scattering coefficient is to the true value.

[0057] When the angular interval between targets exceeds the Rayleigh limit, the imaging results of the proposed method, the real aperture scanning imaging results, and the microwave correlation imaging results obtained by the L1 norm spectral projection gradient algorithm (SPGL1 algorithm) are compared. For example, Figure 6A shows a target scene with an angular interval of 4° between different targets, and Figure 6A contains four targets, with each black dot representing one target. Figures 6B to 6D are schematic diagrams of the imaging results of these three imaging methods in the target scene shown in Figure 6A. The vertical axis of Figures 6A to 6D represents the elevation angle, and the horizontal axis represents the azimuth angle. Specifically, Figure 6B is a schematic diagram of the real aperture scanning imaging result, Figure 6C is a schematic diagram of the SPGL1 algorithm imaging result, and Figure 6D is a schematic diagram of the proposed method imaging result. Table 3 shows the MSE values ​​of these three imaging methods when the angular interval between different targets is 4°.

[0058] When the angular interval between targets is less than the Rayleigh limit, the imaging results of the proposed method, the real aperture scanning imaging results, and the microwave correlation imaging results obtained by the SPGL1 algorithm are further compared. For example, Figure 7A is a target scene diagram when the angular interval between different targets is 2°, and Figure 7A also contains 4 targets, with each black dot representing one target; Figures 7B to 7D are schematic diagrams of the imaging results of these three imaging methods in the target scene shown in Figure 7A. The vertical axis of Figures 7A to 7D represents the elevation angle, and the horizontal axis represents the azimuth angle. Specifically, Figure 7B is a schematic diagram of the real aperture scanning imaging result, Figure 7C is a schematic diagram of the imaging result of the SPGL1 algorithm, and Figure 7D is a schematic diagram of the imaging result of the proposed method. Table 3 shows the MSE values ​​of these three imaging methods when the angular interval between different targets is 2°. Furthermore, for example, Figure 8A is a target scene diagram when the angular interval between different targets is 1°, and Figure 8A also contains 4 targets, with each black dot representing one target; Figures 8B to 8D are schematic diagrams of the imaging results of these three imaging methods in the target scene shown in Figure 8A. The vertical axis of Figures 8A to 8D is the elevation angle, and the horizontal axis is the azimuth angle. Specifically, Figure 8B is a schematic diagram of the real aperture scanning imaging result, Figure 8C is a schematic diagram of the imaging result of the SPGL1 algorithm, and Figure 8D is a schematic diagram of the imaging result of the method proposed in this invention. Table 3 shows the MSE values ​​of these three imaging methods when the angular interval between different targets is 1°.

[0059] Table 3

[0060] When the angular interval between targets is less than the Rayleigh limit, for example, when the angular interval between different targets is 2°, the imaging results of the method proposed in this invention are compared with the microwave correlation imaging results obtained by the SPGL1 algorithm under different signal-to-noise ratios (SNR). The imaging results under different SNRs are shown in Figures 9A to 9F, and the MSE of the super-resolution imaging results under different SNRs is shown in Table 4. In Figures 9A to 9F, the vertical axis represents the elevation angle, and the horizontal axis represents the azimuth angle. Specifically, Figure 9A is a schematic diagram of the imaging result of the SPGL1 algorithm when the angular interval between different targets is 2° and the SNR is 5dB; Figure 9B is a schematic diagram of the imaging result of the method proposed in this invention when the angular interval between different targets is 2° and the SNR is 5dB; Figure 9C is a schematic diagram of the imaging result of the SPGL1 algorithm when the angular interval between different targets is 2° and the SNR is -5dB; Figure 9D is a schematic diagram of the imaging result of the method proposed in this invention when the angular interval between different targets is 2° and the SNR is -5dB; Figure 9E is a schematic diagram of the imaging result of the SPGL1 algorithm when the angular interval between different targets is 2° and the SNR is -15dB; Figure 9F is a schematic diagram of the imaging result of the method proposed in this invention when the angular interval between different targets is 2° and the SNR is -15dB.

[0061] Table 4

[0062] As can be seen from Figures 4A, 4B, 5A, and 5B above, in the random radiation field generated by the joint modulation of frequency and phase in this invention, the radiation fields of the array elements in the sparse array are no longer coherently synthesized in space, resulting in the far-field radiation field of the entire phased array radar exhibiting random fluctuation characteristics. Furthermore, the random radiation field is different at different times, increasing the randomness and flexibility of the random radiation field.

[0063] As can be seen from Figures 6A-6D to 8A-8D and Table 3, as the target spacing decreases, the real aperture scanning imaging method cannot effectively distinguish targets with a target spacing smaller than the Rayleigh limit. The SPGL1 algorithm can effectively distinguish targets and obtain good imaging results when the target spacing is 2°, but when the target spacing continues to decrease, the algorithm cannot obtain good images, and its super-resolution performance is very limited. In contrast, the microwave correlation imaging method based on frequency-phase joint modulation proposed in this invention can effectively utilize echo information, break through the system Rayleigh limit, and achieve super-resolution imaging.

[0064] As shown in Figures 9A-9F and Table 4, noise has a significant impact on microwave correlation imaging using the SPGL1 algorithm. As the signal-to-noise ratio (SNR) decreases, the image reconstruction performance deteriorates, resulting in a large discrepancy between the obtained spatial angular position and the ideal position, leading to false target locations and severely impacting radar detection and identification capabilities. In contrast, the microwave correlation imaging method based on frequency-phase joint modulation proposed in this invention can accurately estimate the target's sparse characteristic parameters and noise feature parameters during the alternating estimation process, better distinguishing the influence of signal and noise, improving target imaging resolution, and reducing noise interference during the optimization process. It can achieve high-performance super-resolution imaging results even at low SNR.

[0065] It should be noted that the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Therefore, features defined as "first" or "second" may explicitly or implicitly include one or more features. In the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0066] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the present invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Furthermore, those skilled in the art can combine and integrate the different embodiments or examples described in this specification.

[0067] In this specification, the word "comprising" does not exclude other components or steps, and "a" or "an" does not exclude multiple instances. While different embodiments may describe certain measures, this does not mean that these measures cannot be combined to produce a good effect.

[0068] The above description, in conjunction with specific preferred embodiments, provides a further detailed explanation of the present invention. It should not be construed that the specific implementation of the present invention is limited to these descriptions. For those skilled in the art, various simple deductions or substitutions can be made without departing from the concept of the present invention, and all such modifications and substitutions should be considered within the scope of protection of the present invention.

Claims

1. A microwave correlation imaging method based on frequency-phase joint modulation, characterized in that, include: Obtain radar parameters; The target imaging scene is divided into grids to obtain multiple spatial angles. Based on the radar parameters, the multiple spatial angles, the frequency modulation function, and the phase modulation function, a random radiation field with joint frequency and phase modulation is constructed. The target echo signal is received using the random radiation field with joint frequency and phase modulation, and the received echo signal is subjected to range compression processing to obtain a range pulse compression envelope. Based on the random radiation field with joint frequency and phase modulation, the range pulse compression envelope, a sub-optimization problem of noise variance, a sub-optimization problem of target sparse characteristic parameters, and a sub-optimization problem of super-resolution image reconstruction combining the noise variance and the target sparse characteristic parameters, a super-resolution image of the target that meets preset requirements is determined by alternately estimating the noise variance, the target sparse characteristic parameters, and the super-resolution image.

2. The microwave correlation imaging method based on frequency-phase joint modulation according to claim 1, characterized in that, The radar parameters include: radar electromagnetic parameters, phased array structural parameters, and radar pulse repetition count. The phased array structural parameters include: the total number of array elements and the arrangement position of each array element.

3. The microwave correlation imaging method based on frequency-phase joint modulation according to claim 2, characterized in that, The step of constructing a frequency-phase jointly modulated random radiation field based on the radar parameters, the multiple spatial angles, the frequency modulation function, and the phase modulation function includes: constructing a random radiation field matrix of frequency-phase jointly modulated by different pulses based on the radar electromagnetic parameters, the arrangement position of each array element, the number of radar pulse repetitions, the multiple spatial angles, the antenna beam pointing, the frequency modulation function, and the phase modulation function. The first dimension of the random radiation field matrix of frequency-phase jointly modulated by different pulses is the number of radar pulse repetitions, and the second dimension is the total number of spatial angles. Furthermore, the random radiation field matrix of frequency-phase jointly modulated by different pulses represents the random radiation field of frequency-phase jointly modulated by the radar.

4. The microwave correlation imaging method based on frequency-phase joint modulation according to claim 1 or 3, characterized in that, At least one of the frequency modulation function and the phase modulation function is an independent random process function that varies with time. The independent random process function that varies with time refers to a function that takes time as the independent variable, has a corresponding random value at each time point, and the changes in the random values ​​corresponding to different and non-overlapping time intervals are independent of each other.

5. The microwave correlation imaging method based on frequency-phase joint modulation according to claim 3, characterized in that, The expression for any element in the random radiation field matrix jointly modulated by the frequencies and phases of the different pulses is as follows: ;in, Represents any one of the elements. 、 、 All are positive integers. ， ， ， This represents the total number of spatial angles. This indicates the number of times the radar pulse is repeated. This represents the total number of elements in a phased array. and The first The azimuth and elevation angles corresponding to each spatial angle This refers to the pointing position of the antenna beam, and and These represent the azimuth and elevation angles, respectively. Indicates the first The arrangement of each array element, and and The first The x and y coordinates of each array element in the coordinate system Indicates the first pulse At the time The amplitude of the excitation current on each array element and They represent the first pulse At the time The modulation frequency and modulation phase added to each array element The wavelength is represented in the radar electromagnetic parameters. Represents the speed of light. The imaginary unit, Represents the sine function. Represents the cosine function. This represents the natural exponential function.

6. The microwave correlation imaging method based on frequency-phase joint modulation according to claim 1, characterized in that, The sub-optimization problems based on the random radiation field of the frequency-phase joint modulation, the range pulse compression envelope, the noise variance, the target sparse characteristic parameters, and the super-resolution image reconstruction problem combining the noise variance and the target sparse characteristic parameters, determine the super-resolution image of the target that meets preset requirements by alternately estimating the noise variance, the target sparse characteristic parameters, and the super-resolution image. This includes: in the first step of estimating the noise variance, the target sparse characteristic parameters, and the super-resolution image... When performing the alternating estimation, obtain the first... The super-resolution image estimated in the first estimation, based on the first... The first estimated super-resolution image, the sub-optimization problem of the noise variance, the random radiation field of the frequency-phase joint modulation, and the range pulse compression envelope determine the second... The noise variance estimated in the first estimation, and according to the first... The sub-optimization problem of the estimated super-resolution image and the target sparse property parameters determines the first... The target sparse property parameters are estimated in the second step; It is a positive integer greater than or equal to 1, when When =1, the super-resolution image estimated for the 0th time is the pre-generated initial super-resolution image; based on the frequency-phase jointly modulated random radiation field, the range pulse compression envelope, and the first... The noise variance of the first estimate, the first The target sparse property parameters estimated in the first estimation, the sub-optimization problem of super-resolution image reconstruction, and the determination of the first using an optimization algorithm. The super-resolution image estimated in the second step; determine the first... If the estimated super-resolution image meets the preset requirements, then the first estimated super-resolution image is... The estimated super-resolution image is taken as the super-resolution image of the target that meets the preset requirements; if it does not meet the requirements, then let Then, continue to analyze the noise variance, the target sparsity parameter, and the first super-resolution image. The estimation is repeated several times until a super-resolution image of the target that meets the preset requirements is obtained.

7. The microwave correlation imaging method based on frequency-phase joint modulation according to claim 6, characterized in that, The random radiation field modulated according to the frequency and phase, the range pulse compression envelope, and the first The noise variance of the first estimate, the first The target sparse property parameters estimated in the first estimation, the sub-optimization problem of super-resolution image reconstruction, and the determination of the first using an optimization algorithm. The super-resolution image estimated in the second step includes: a random radiation field jointly modulated by the frequency and phase, the range pulse compression envelope, and the first... The noise variance of the first estimate, the first The estimated target sparse characteristic parameters are substituted into the sub-optimization problem of super-resolution image reconstruction, and then the derivative of the sub-optimization problem is obtained to yield the equation to be solved. An optimization algorithm is then used to iteratively solve the equation to obtain the... Super-resolution image estimated at the second time.

8. The microwave correlation imaging method based on frequency-phase joint modulation according to claim 6, characterized in that, The determination of the first Whether the super-resolution image estimated in the first step meets the preset requirements includes: determining whether the first... The super-resolution image estimated in the second step and the first step The difference between the super-resolution images estimated in the first step; the difference is determined based on the relationship between the difference and a preset threshold. Whether the super-resolution image estimated in the second step meets the preset requirements, wherein when the difference is less than or equal to the preset threshold, it indicates that the second step... The estimated super-resolution image meets the preset requirements. When the difference is greater than the preset threshold, it indicates that the first estimated super-resolution image meets the preset requirements. The estimated super-resolution image does not meet the preset requirements.

9. The microwave correlation imaging method based on frequency-phase joint modulation according to claim 1, characterized in that, The method for constructing the sub-optimization problems of the noise variance, the target sparse characteristic parameters, and the super-resolution image reconstruction includes: constructing a microwave correlation imaging model of the frequency-phase jointly modulated random radiation field based on the frequency-phase jointly modulated random radiation field, the range pulse compression envelope, and the phase matrix, wherein the microwave correlation imaging model includes a noise vector and a super-resolution image to be reconstructed; transforming the microwave correlation imaging model into a multi-parameter optimization problem under sparse constraints based on maximum a posteriori estimation and maximum likelihood estimation criteria, wherein the multi-parameter optimization problem includes the super-resolution image to be reconstructed, the noise variance to be solved, and the target sparse characteristic parameters to be solved; and decomposing the multi-parameter optimization problem into the sub-optimization problems of the noise variance, the target sparse characteristic parameters, and the super-resolution image reconstruction.

10. A microwave correlation imaging device based on frequency-phase joint modulation, comprising a processor, a communication interface, a memory, and a communication bus, characterized in that, The processor, the communication interface, and the memory communicate with each other via the communication bus; the memory is used to store computer programs; the processor, when executing the program stored in the memory, implements the steps of the method described in any one of claims 1-9.

Images