A curved array implicit aperture extension imaging method based on sparse enhanced iterative adaptive processing

Abstract

The application discloses a curved surface array implicit aperture expansion imaging method based on sparse enhanced iterative adaptive processing, belongs to the technical field of array signal processing and beam imaging, and particularly relates to a high-resolution imaging algorithm for a non-planar (curved surface) conformal array. The core realizes accurate beam focusing, sidelobe suppression and robust beam pointing through virtual element expansion and iterative adaptive (IAA) beam shaping technology, and can be widely adapted to various systems such as radars, sonars and radio positioning systems.

Description

Technical Field

[0001] This invention belongs to the field of array signal processing and beam imaging technology, specifically, it relates to a surface array implicit aperture extension imaging method based on sparse enhancement iterative adaptive processing. Background Technology

[0002] Beamforming and array signal processing are core supporting technologies for achieving efficient detection, accurate positioning, and interference suppression in systems such as vehicle-mounted millimeter-wave radar and aircraft skin radar. Signal processing schemes for conformal surface arrays have become a research hotspot in this field. Conformal surface array imaging and beamforming optimize beam pointing accuracy, spatial resolution, and anti-interference capabilities by adapting signal modeling and processing logic to the geometric configuration of the curved surface. Its performance directly affects the system's target detection reliability and environmental adaptability in complex scenarios.

[0003] Although signal processing for conformal arrays has been extensively studied, with the development of fields such as intelligent connected vehicles and aerospace sensing, traditional processing methods face greater challenges in terms of performance balance and engineering robustness. The geometric characteristics of conformal arrays determine that their signal processing has the following inherent limitations:

[0004] It is not adaptable to scanning angles and environmental complexity. Traditional planar hard projection processing is prone to phase distortion when scanning at large angles. When facing complex environments such as multipath coherent interference and urban clutter, the stability of signal resolution and weak target extraction is difficult to guarantee.

[0005] The flexibility of applicable scenarios is limited. Most solutions rely on ideal array models or specific curvature assumptions, which are not robust to engineering disturbances such as array installation deviations and mutual coupling of array elements, making it difficult to quickly adapt to the dynamic requirements of different carrier configurations (such as vehicle bumpers and aircraft skins).

[0006] Currently, numerous studies have proposed relatively comprehensive solutions for signal processing of conformal surface arrays. Some studies have proposed simplified processing algorithms based on the hard projection of the surface as a plane, achieving beamforming by applying mature planar array algorithms. However, these methods, based on geometric approximation assumptions, are prone to truncation errors and phase distortions during large-view scanning, failing to accurately reproduce the true manifold characteristics of the surface array. Other studies have introduced super-resolution algorithms (such as MUSIC) to improve imaging resolution, but these methods are prone to performance degradation in multipath coherent scenarios due to covariance matrix rank deficiency, resulting in insufficient suppression of coherent interference. Furthermore, some studies have employed iterative adaptive algorithms (such as the standard IAA) to optimize the spatial spectrum, but have not fully discussed the impact of the nonlinear phase gradient of the surface on algorithm convergence and sidelobe suppression, leaving room for improvement in the detection performance of weak targets in complex clutter environments.

[0007] Therefore, existing surface conformal array signal processing schemes often suffer from difficulties in simultaneously achieving super-resolution imaging and wide-angle scanning accuracy, as well as insufficient robustness under coherent interference environments, making them unsuitable for the complex detection requirements of real-world engineering scenarios. An optimized scheme is needed that organically integrates the concept of virtual array element expansion, direct surface manifold modeling, and super-resolution adaptive processing, aiming to achieve super-resolution imaging, large-field-of-view distortion-free scanning, and anti-coherent interference capabilities without increasing hardware costs, thus better adapting to the application needs of practical engineering environments such as automotive and aerospace.

[0008] In view of this, the present invention is proposed. Summary of the Invention

[0009] To solve the above-mentioned technical problems, the basic concept of the technical solution adopted by the present invention is as follows:

[0010] Given the application requirements of conformal curved arrays in scenarios such as vehicle-mounted millimeter-wave radar and aerospace sensing, and the limitations of existing processing methods such as phase distortion during large-view scanning, insufficient coherent source resolution, and spatial resolution limited by physical aperture, we attempt to explore a virtual aperture expansion imaging method for curved arrays based on the Sparsity-Enhanced Iterative Adaptive Approach (SE-IAA). This method aims to improve the spatial resolution, large-view scanning accuracy, and anti-coherent interference capability of curved arrays without increasing the number of physical array elements through a fusion architecture of "in-situ physical modeling + implicit aperture expansion," hoping to provide a feasible optimization solution for complex detection scenarios. The specific technical solution adopted in this invention is as follows:

[0011] The proposed method for virtual aperture extension imaging based on SE-IAA surface arrays includes the following steps:

[0012] Step 1: Constructing the in-situ high-resolution manifold dictionary matrix for the surface: Obtain the three-dimensional physical coordinates of each element of the surface array, and combine the electromagnetic wave propagation characteristics to construct a guide vector dictionary matrix that can accurately reflect the manifold characteristics of the surface array for the preset full-space scanning grid, providing a high-precision benchmark model for subsequent iterative processing.

[0013] Step 2: Initial estimation of spatial power spectrum: Acquire multi-channel echo signals received by the curved array, calculate the sample covariance matrix, and use conventional beamforming (DAS) algorithm to obtain the initial spatial power spectrum, thus initially locking the approximate azimuth region where the target is located;

[0014] Step 3: Virtual covariance matrix reconstruction: Using the current power spectrum estimate combined with the in-situ dictionary matrix, the idealized model covariance matrix is ​​reconstructed. The information gaps between physical sampling points are "filled" through mathematical logic, which effectively simulates the correlation characteristics of a larger aperture virtual array. At the same time, a dynamic diagonal loading term is introduced to ensure the numerical stability of matrix inversion.

[0015] Step 4: Nonlinear sparse sharpening: Perform nonlinear power operation on the iteratively updated power spectrum to construct a nonlinear weighting field that "strengthens and suppresses weakness", which promotes the spatial spectrum to converge towards the discrete pulse shape, effectively suppressing sidelobe interference and compressing the main lobe width.

[0016] Step 5: Iterative convergence judgment and result output: Steps 3 and 4 are executed alternately. When the change in power spectrum between two adjacent iterations is lower than the preset threshold, the iteration is stopped and the final spatial spectrum imaging result is output.

[0017] Compared with the prior art, the present invention has the following advantages:

[0018] By using virtual aperture implicit expansion, phase lossless correction, and coherent signal decoupling processing, it breaks through the resolution bottleneck of traditional curved arrays, eliminates large-angle scanning distortion, and enhances signal resolution robustness and positioning accuracy in complex environments, making it more suitable for detection scenarios with multipath interference and large field of view scanning. Specifically, the algorithm achieves technological breakthroughs through multi-module collaboration: First, by using the "implicit aperture expansion" logic of virtual array element expansion, the observation degrees of freedom are expanded and the beam main lobe width is compressed without changing the physical size of the antenna and the number of array elements. Second, by integrating the idea of ​​virtual array element expansion, the "implicit expansion" logic of virtual array elements is used to assist in the direct modeling of the curved array manifold, replacing the lossy transformation of traditional hard projection of curved surfaces into planes. Through lossless phase correction and implicit supplementation of virtual aperture, the phase distortion and positioning deviation problems in large-angle and edge field-of-view scanning are alleviated, improving imaging accuracy. Third, the iterative adaptive (IAA) algorithm is introduced to decoherently process the covariance matrix, overcoming the rank deficiency of traditional super-resolution algorithms (such as MUSIC) in multipath coherent scenarios, and achieving accurate resolution of coherent signal sources. At the same time, the nonlinear adjustment mechanism deeply suppresses spatial spectrum sidelobes and background noise, improving the ability to extract weak targets in complex clutter environments.

[0019] The specific embodiments of the present invention will now be described in further detail with reference to the accompanying drawings. Attached Figure Description

[0020] Figure 1 This is a flowchart illustrating the overall implementation process of the algorithm.

[0021] Figure 2 This is a performance comparison chart of the "spatial power spectrum" of the present invention and traditional beamforming algorithms under different signal-to-noise ratios;

[0022] Figure 3 This is a performance comparison chart of the "two-dimensional spatial position (xy) imaging results" of the present invention and traditional beamforming algorithms;

[0023] Figure 4 This is a performance comparison chart of the spatial spectrum of the present invention and traditional beamforming algorithms in a coherent source scenario.

[0024] Figure 5 This is a performance comparison chart of the "spatial spectrum comparison chart" between the present invention and traditional beamforming algorithms in scenarios where independent and coherent sources coexist.

[0025] Figure 6 This is a performance comparison chart of the spatial spectrum of the present invention and traditional beamforming algorithms under different snapshot numbers. Detailed Implementation

[0026] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments will be clearly and completely described below with reference to the accompanying drawings. The following embodiments are used to illustrate the present invention.

[0027] This embodiment first establishes the physical topology model of the curved array. For a typical cylindrical conformal array, its elements are distributed on a non-planar curved surface carrier. To avoid phase distortion caused by planar mapping, this invention directly utilizes the precise spatial physical coordinates of the elements within the horizontal observation plane. Modeling is performed.

[0028] Let the radius of the cylindrical array be... , No. The corner positions of the physical array elements are Then its spatial coordinates are precisely represented as:

[0029]

[0030] Unlike traditional uniform linear arrays (ULA), this invention fully preserves the depth direction when constructing the steering vector. The phase offset of the axis. For any scanning angle in space. Its corresponding in-situ steering vector The Each component is defined as:

[0031]

[0032] in, The wavelength of the signal is denoted as λ. The establishment of this model ensures that the algorithm can restore the original phase relationship of the electromagnetic wave with the highest fidelity across the entire space (especially in the large-view scanning area), laying the accuracy foundation for subsequent implicit extensions.

[0033] The implementation steps of this invention can be divided into four stages: initial perception, matrix reconstruction, nonlinear sharpening, and convergent output. The specific process is as follows:

[0034] Step 1: Construct a high-resolution in-situ dictionary matrix

[0035] Based on the aforementioned physical coordinates, a pre-defined full-space scanning grid is used. Above, construct a high-resolution guided vector dictionary matrix. :

[0036]

[0037] This dictionary matrix serves as the basis for subsequent iterative processing, through dense grid partitioning (such as...). (Stepping) ensures precise characterization of the signal source location.

[0038] Step 2: Initial estimation of spatial power spectrum

[0039] Acquire the received original surface multi-channel signal Calculate the sample covariance matrix A preliminary spatial scan was performed using a conventional beamforming (DAS) algorithm to obtain the initial power spectrum distribution. This step aims to initially pinpoint the approximate location of the target.

[0040] Step 3: Virtual Covariance Matrix Filling and Reconstruction. This is a key step in realizing "implicit virtual aperture expansion." This invention no longer directly processes the noisy and rank-deficient sample matrix, but instead utilizes the current power spectrum estimate. In conjunction with the in-situ dictionary Reconstruct an idealized model covariance matrix :

[0041]

[0042] in, This is a dynamic diagonal loading term used to ensure the numerical stability of matrix inversion. Physically, The matrix "fills" the information vacuum between physical sampling points through mathematical logic, and the amount of information it contains is equivalent to the correlation features that a virtual array with a larger physical size can provide.

[0043] Step 4: Nonlinear sparse sharpening processing

[0044] In obtaining the updated power spectrum Subsequently, a nonlinear sparse sharpening operator is introduced to nonlinearly intervene in the iterative process. Its mathematical implementation involves performing a Hadamard exponentiation operation on the power spectral vector:

[0045]

[0046] In a preferred embodiment, take (i.e., execution) (Sharpening logic). This operator forcibly compresses the width of the main lobe beam by constructing a nonlinear weighting field that "strengthens and weakens" the main lobe. From the theoretical perspective of Sparse Recovery, this operation is similar to introducing a sharpening logic into the iteration. The norm penalty term causes the spatial spectrum to converge toward a discrete pulse shape (Dirac-delta distribution), thereby completely eliminating sidelobe interference.

[0047] Step 5: Iterative convergence judgment and result output

[0048] The algorithm executes steps S3 and S4 alternately, through The observation space is adaptively "whitened". When the change in power spectrum between two consecutive iterations is lower than a preset threshold, the algorithm stops iterating and outputs the final super-resolution spatial spectrum image.

[0049] To objectively demonstrate the improved effects of this invention, the following description is based on specific experimental data:

[0050] (1) Verification of spatial resolution (corresponding appendix) Figure 2 ):exist In a coherent source scenario, the spatial spectra of various algorithms are compared. As shown in the figure, conventional beamforming (DAS) exhibits a broad, blunt single peak; the traditional subspace algorithm (MUSIC) completely fails due to the rank deficiency caused by the coherent source. In contrast, the proposed (SE-IAA) algorithm generates two extremely sharp spectral lines, accurately separating the coherent source. and The coherent target. Its main lobe width is compressed by about 40% compared to the standard algorithm.

[0051] (2) Positioning accuracy analysis (corresponding appendix) Figure 3 Stability is demonstrated through RMSE vs. SNR curves. Experimental data show that this invention performs well at low signal-to-noise ratios (e.g., ...). It can still maintain extremely high beam pointing accuracy, and the RMSE converges rapidly to [value missing]. Within this range, it demonstrates robustness similar to SPICE and superior to the standard IAA.

[0052] (3) Computational performance evaluation (corresponding appendix) Figure 4 This invention provides the highest imaging resolution while achieving faster single-run time than similar sparse algorithms (such as SPICE) through efficient iterative logic control, demonstrating its application potential in real-time processing systems for practical engineering.

[0053] This invention belongs to the field of array signal processing and beamforming technology, specifically relating to a high-resolution imaging algorithm for non-planar (curved) conformal arrays. The core of the algorithm utilizes virtual element expansion and iterative adaptive (IAA) beamforming technology to achieve precise beam focusing, sidelobe suppression, and robust beam pointing, making it widely adaptable to various systems such as radar, sonar, and radio positioning.

[0054] Typical engineering applications of this technology include:

[0055] In the field of radar detection: supporting autonomous driving environmental perception for vehicle-mounted millimeter-wave radar and omnidirectional target detection for aircraft skin phased array radar;

[0056] In the field of underwater acoustics and communications: underwater target identification adapted to conformal sonar, and signal source tracking for radio positioning systems;

[0057] Specialized applications in complex environments: Especially suitable for detection scenarios with stringent spatial resolution requirements and multipath coherent interference, it can effectively improve the positioning accuracy of radio frequency / acoustic signals, the anti-distortion capability of large-angle scanning, and the robustness against clutter.

Claims

1. A method for implicit aperture extension imaging of curved arrays based on sparse enhancement-type iterative adaptive processing, characterized in that, Includes the following steps: Step 1: Constructing the in-situ high-resolution manifold dictionary matrix for the surface: Obtain the three-dimensional physical coordinates of each element of the surface array, and combine the electromagnetic wave propagation characteristics to construct a guide vector dictionary matrix that can accurately reflect the manifold characteristics of the surface array for the preset full-space scanning grid, providing a high-precision benchmark model for subsequent iterative processing. Step 2: Initial estimation of spatial power spectrum: Acquire multi-channel echo signals received by the curved array, calculate the sample covariance matrix, and use conventional beamforming (DAS) algorithm to obtain the initial spatial power spectrum, thus initially locking the approximate azimuth region where the target is located; Step 3: Virtual covariance matrix reconstruction: Using the current power spectrum estimate combined with the in-situ dictionary matrix, the idealized model covariance matrix is ​​reconstructed. The information gaps between physical sampling points are "filled" through mathematical logic, which effectively simulates the correlation characteristics of a larger aperture virtual array. At the same time, a dynamic diagonal loading term is introduced to ensure the numerical stability of matrix inversion. Step 4: Nonlinear sparse sharpening: Perform nonlinear power operation on the iteratively updated power spectrum to construct a nonlinear weighting field that "strengthens and suppresses weakness", which promotes the spatial spectrum to converge towards the discrete pulse shape, effectively suppressing sidelobe interference and compressing the main lobe width. Step 5: Iterative convergence judgment and result output: Steps 3 and 4 are executed alternately. When the change in power spectrum between two adjacent iterations is lower than the preset threshold, the iteration is stopped and the final spatial spectrum imaging result is output.

2. The implicit aperture extension imaging method for curved arrays based on sparse enhancement iterative adaptive processing according to claim 1, characterized in that, The curved array is a cylindrical conformal array, with its elements distributed on a non-planar curved surface carrier. It utilizes the precise spatial physical coordinates of the elements within the horizontal observation plane. Modeling is performed, and the radius of the cylindrical array is set to... , No. The corner positions of the physical array elements are Then its spatial coordinates are precisely represented as: 。 3. The implicit aperture extension imaging method for curved arrays based on sparse enhancement iterative adaptive processing according to claim 2, characterized in that, When constructing the guide vector dictionary matrix in step 1, for any scanning angle in space... Its corresponding in-situ steering vector The Each component is defined as: 。 in, The wavelength is the signal wavelength.

4. The implicit aperture extension imaging method for surface arrays based on sparse enhancement iterative adaptive processing according to claim 1, characterized in that, The detailed algorithm for the initial estimation of the spatial power spectrum in step 2 includes: acquiring the received original surface multi-channel signal. Calculate the sample covariance matrix A preliminary spatial scan was performed using a conventional beamforming (DAS) algorithm to obtain the initial power spectrum distribution. .

5. The implicit aperture extension imaging method for curved arrays based on sparse enhancement iterative adaptive processing according to claim 1, characterized in that, The detailed algorithm for reconstructing the virtual covariance matrix in step 3 includes: using the current power spectrum estimate... In conjunction with the in-situ dictionary Reconstruct an idealized model covariance matrix : 。 in, This is a dynamic diagonal loading term used to ensure the numerical stability of matrix inversion.

6. The implicit aperture extension imaging method for surface arrays based on sparse enhancement iterative adaptive processing according to claim 1, characterized in that, The specific computational method for nonlinear sparse sharpening in step 4 includes: obtaining the updated power spectrum. Subsequently, a nonlinear sparse sharpening operator is introduced to nonlinearly intervene in the iterative process. Its mathematical implementation involves performing a Hadamard exponentiation operation on the power spectral vector: 。 7. The implicit aperture extension imaging method for curved arrays based on sparse enhancement iterative adaptive processing according to claim 1, characterized in that, In step 5, the algorithm alternately executes steps S3 and S4, through... The observation space is adaptively "whitened". When the change in power spectrum between two consecutive iterations is lower than a preset threshold, the algorithm stops iterating and outputs the final super-resolution spatial spectrum image.

Images