Compressive sensing based sparse-aperture radar high-resolution imaging method and system

Abstract

This application provides a high-resolution imaging method and system for sparse aperture radar based on compressed sensing, relating to the field of radar imaging technology. This application selects target transmitting and receiving elements based on target sparsity, the spatial distribution of radar elements, and the coherence constraints of candidate antenna subarrays to determine the working transmitting and receiving channels. The working transmitting channel is controlled to transmit frequency-agile signals in a frequency-hopping sequence, and the working receiving channel receives the echoes and binds the sampling position and frequency data to form sparse echo data. A two-dimensional phase constraint is established based on the phase breakpoints in the spatial sampling dimension and the frequency-hopping sequence to obtain the phase trajectory. The target plane coordinates are determined using the intersection information of the phase trajectory on the two-dimensional imaging plane, and a high-resolution image is generated by combining the sparse echo data. This approach can improve target localization and imaging stability while reducing channel and sampling burden.

Description

Technical Field

[0001] This application relates to the field of radar imaging technology, and in particular to a high-resolution imaging method and system for sparse aperture radar based on compressed sensing. Background Technology

[0002] Sparse aperture radar high-resolution imaging is a technology for imaging target scenes with radar when antenna array elements, sampling positions, or observation data are incomplete. It is applicable to radar platforms such as vehicle-mounted, airborne, and spaceborne radars, and has promising applications in scenarios such as road monitoring, ground object identification, and target detection.

[0003] Existing technology discloses a compressed sensing imaging method for synthetic aperture radar. It constructs a standard orthogonal basis for radar echo signals through the ChirpScaling operator and performs two-dimensional sparsity optimization in the range and azimuth directions using compressed sensing to obtain high-resolution radar images with a low data volume. This reduces the pressure on data storage and transmission and is an existing solution in the field of compressed sensing radar imaging.

[0004] However, under conditions of sparse aperture, discontinuous sampling, or multi-frequency observation, phase changes in echo data are more prone to reversal and jumps. Relying solely on conventional two-dimensional sparse optimization processes makes it difficult to stably determine the continuous phase relationship corresponding to the target scattering point. When the phase relationship recovery is unstable, both target position calculation and image intensity representation may be affected. Therefore, existing technologies suffer from insufficient stability in target phase recovery and high-resolution imaging under sparse sampling conditions. Summary of the Invention

[0005] The purpose of this application is to provide a high-resolution imaging method and system for sparse aperture radar based on compressed sensing, so as to solve the problems of insufficient target phase recovery and high-resolution imaging stability under sparse sampling conditions in the prior art.

[0006] To address the aforementioned technical problems, in a first aspect, this application provides a high-resolution imaging method for sparse aperture radar based on compressed sensing, the method comprising: Based on the target sparsity of the target scene, the spatial distribution of radar elements in the antenna array, and the coherence constraints of the candidate antenna subarrays, target transmitting elements and target receiving elements participating in the operation are selected in the antenna array to select the corresponding operating transmitting channel and operating receiving channel. The working transmitting channel is controlled to sequentially transmit frequency agile signals in a preset frequency switching order at multiple spatial sampling points, and the working receiving channel receives the echo signals reflected from the target scene, records the sampling position and frequency data corresponding to the echo signals, and obtains sparse echo data. Based on the first phase breakpoint of the spatial sampling dimension determined by the sampling position and the second phase breakpoint of the frequency jump sequence, a two-dimensional phase constraint is constructed, and the phase periodicity characteristics of the two-dimensional phase constraint are used to unwrap the target scattering point to obtain the first phase trajectory and the second phase trajectory. Using the intersection information of the first phase trajectory and the second phase trajectory in the two-dimensional imaging plane, the target plane coordinates of the target scattering point are calculated, and a high-resolution image is generated based on the sparse echo data and the target plane coordinates.

[0007] Optionally, the construction of a two-dimensional phase constraint based on the first phase breakpoint of the spatial sampling dimension determined by the sampling position and the second phase breakpoint of the frequency transition sequence includes: The echo signal of the sparse echo data within the same distance cell is extracted, and the wrapped phase of the echo signal is retained to obtain the measurement phase; The phase change of the measured phase between adjacent sampling positions is calculated along the spatial sampling dimension to generate a first phase residual, and the phase change of the measured phase between adjacent frequency data is calculated along the frequency jump sequence to generate a second phase residual. The phase jump value generated by the non-uniform sampling interval of the spatial sampling dimension in the first phase residual is determined as the first phase breakpoint, and the phase jump value generated by the frequency jump sequence in the second phase residual is determined as the second phase breakpoint. Based on the distribution positions of the first phase breakpoint and the second phase breakpoint in different dimensions, a two-dimensional phase constraint is established for performing phase unwrapping.

[0008] Optionally, the step of calculating the phase change of the measured phase between adjacent sampling positions along the spatial sampling dimension to generate a first phase residual, and calculating the phase change of the measured phase between adjacent frequency data along the frequency jump sequence to generate a second phase residual, includes: The spatial sampling interval between adjacent sampling positions and the frequency hopping step size between adjacent frequency data are obtained respectively. A first baseline phase and a second baseline phase are constructed based on the spatial sampling interval and the frequency hopping step size. The difference value of the measured phase between adjacent sampling positions is calculated along the spatial sampling dimension, and the first baseline phase is used to perform phase compensation on the difference value to obtain the first phase difference value; The difference value of the measured phase between adjacent frequency data is calculated along the frequency jump sequence, and the second baseline phase is used to perform phase compensation on the difference value to obtain the second phase difference value; Modulo operations are performed on the first phase difference and the second phase difference respectively, and the operation results are constrained within the corresponding preset intervals to generate the first phase residual and the second phase residual.

[0009] Optionally, the step of using the phase periodicity characteristic of the two-dimensional phase constraint to perform phase unwrapping on the target scattering point to obtain the first phase trajectory and the second phase trajectory includes: The phase period characteristics are determined based on the phase period between the first phase breakpoint and the second phase breakpoint in the two-dimensional phase constraint. By solving the unwrapping equations corresponding to each phase residual, the phase step lost in the first and second phase residuals is obtained. The unwrapping equations are determined based on the phase periodicity characteristics. The phase step is superimposed and compensated to the first phase residual and the second phase residual respectively, to reconstruct the continuously monotonically changing first phase trajectory and second phase trajectory.

[0010] Optionally, calculating the target plane coordinates of the target scattering point using the intersection information of the first phase trajectory and the second phase trajectory in the two-dimensional imaging plane includes: The first phase trajectory and the second phase trajectory are projected onto a two-dimensional coordinate grid constructed by the azimuth and range directions, respectively, to generate corresponding spatial geometric curves; Calculate the vertical distance between the spatial geometric curves, and take the physical coordinate point corresponding to the intersection position where the vertical distance reaches the minimum value as the initial plane coordinates; Based on the initial plane coordinates, combined with the spatial sampling points and the frequency transition sequence, the corresponding reference phase is calculated; The reference phase is compared with the measured phase. If the difference is less than or equal to a preset convergence threshold, the initial plane coordinates are determined to correspond to the target scattering point. The initial plane coordinates are confirmed as the target plane coordinates of the target scattering point.

[0011] Optionally, generating a high-resolution image based on the sparse echo data and the target plane coordinates includes: Obtain the initial scattering intensity corresponding to each target plane coordinate in the target plane coordinate set; Based on the spatial distribution of radar elements in the antenna array, the pattern gain at the target plane coordinates is calculated, and the initial scattering intensity is compensated using the pattern gain to calculate the target scattering intensity. In a preset two-dimensional grid, the nodes corresponding to the target plane coordinates are assigned the target scattering intensity to generate a high-resolution image.

[0012] Optionally, the step of calculating the pattern gain at the target plane coordinates based on the spatial distribution of radar elements in the antenna array, and using the pattern gain to perform gain compensation on the initial scattering intensity to calculate the target scattering intensity includes: Calculate the relative spatial distance between each radar element and the plane coordinates of each target based on the spatial distribution of each radar element. Based on the relative spatial distance, calculate the spatial weighting coefficients of the projection of each radar element onto the plane coordinates of each target; The spatial weighting coefficients corresponding to the radar array elements are coherently superimposed to calculate the radiation pattern gain at the target plane coordinates. The reciprocal of the pattern gain is used as the gain compensation coefficient, and the gain compensation coefficient is multiplied by the initial scattering intensity to calculate the target scattering intensity.

[0013] Optionally, controlling the working transmitting channel to sequentially transmit frequency-agile signals in a preset frequency-hopping order at multiple spatial sampling points, and having the working receiving channel receive the echo signal reflected from the target scene, includes: The working transmission channel is controlled to move along the trajectory of the radar platform and transmit frequency-agile signals sequentially at the multiple spatial sampling points. The carrier frequency of the frequency-agile signal is switched according to the preset frequency jump sequence. The working receiving channel is controlled to synchronously receive the echo signal formed after the frequency agile signal is reflected by the target scene; The spatial coordinates of the echo signal when it is received are extracted as the sampling position, and the carrier frequency of the echo signal is extracted as frequency data. The echo signal is matched and bound with the corresponding sampling position and frequency data to generate sparse echo data.

[0014] Optionally, the step of selecting target transmitting elements and target receiving elements to participate in the operation from the antenna array based on the target sparsity of the target scene, the spatial distribution of radar elements in the antenna array, and the coherence constraints of candidate antenna subarrays includes: Based on the target sparsity of the target scenario, set the upper limit value of the channels that can be opened simultaneously. The spatial pointing characteristics of each radar element in the candidate antenna subarray are calculated based on the spatial distribution location, and the correlation analysis is performed on the spatial pointing characteristics corresponding to different radar elements in the candidate antenna subarray to obtain the cross-correlation coefficient used to characterize the coherence constraint. Within the upper limit of the channel, the target transmitting element and the target receiving element with non-uniform physical spacing are selected from the candidate antenna subarrays by using the minimization of the cross-correlation coefficient as the screening condition.

[0015] Secondly, this application provides a high-resolution imaging system for sparse aperture radar based on compressed sensing, the system comprising: The selection module is used to select target transmitting elements and target receiving elements in the antenna array based on the target sparsity of the target scene, the spatial distribution of radar elements in the antenna array, and the coherence constraints of candidate antenna subarrays, so as to select the corresponding working transmitting channel and working receiving channel. The recording module is used to control the working transmission channel to sequentially transmit frequency agile signals in a preset frequency switching order at multiple spatial sampling points, and the working receiving channel to receive the echo signals reflected by the target scene, record the sampling position and frequency data corresponding to the echo signals, and obtain sparse echo data. The construction module is used to construct a two-dimensional phase constraint based on the first phase breakpoint of the spatial sampling dimension determined by the sampling position and the second phase breakpoint of the frequency jump sequence, and to use the phase periodicity characteristics of the two-dimensional phase constraint to perform phase unwrapping on the target scattering point to obtain the first phase trajectory and the second phase trajectory. The generation module is used to calculate the target plane coordinates of the target scattering point using the intersection information of the first phase trajectory and the second phase trajectory in the two-dimensional imaging plane, and generate a high-resolution image based on the sparse echo data and the target plane coordinates.

[0016] The high-resolution imaging method for sparse aperture radar based on compressed sensing provided in this application has the following advantages: By selecting target transmitting and receiving elements according to the target sparsity of the target scene, the spatial distribution of radar elements in the antenna array, and the coherence constraints of candidate antenna subarrays, redundant channels can be reduced and observation differences can be improved; by controlling the working transmitting channel to transmit frequency-agile signals at multiple spatial sampling points and recording the sampling position and frequency data, the spatial and frequency correspondence of the echo signal can be preserved while reducing the amount of sampling; by constructing a two-dimensional phase constraint based on the first and second phase breakpoints and performing phase unwrapping, phase jump misjudgment can be reduced, making the phase change of the target scattering point more continuous; by using the intersection information of the first and second phase trajectories to calculate the target plane coordinates and generate a high-resolution image, the stability of target scattering point localization and image representation can be improved.

[0017] Furthermore, by extracting sparse echo data within the same distance cell and retaining the entangled phase, the measured phase is obtained; then, the phase change is calculated along the spatial sampling dimension and the frequency jump sequence, respectively, to obtain the first phase residual and the second phase residual; subsequently, the phase jump values ​​caused by the non-uniform sampling interval and the frequency jump sequence are identified to determine the first phase breakpoint and the second phase breakpoint; finally, a two-dimensional phase constraint is established based on the distribution positions of the phase breakpoints in the two dimensions, thereby enabling the use of phase information in both the spatial and frequency directions for phase unwrapping, improving the reliability of continuous phase recovery. Attached Figure Description

[0018] To more clearly illustrate the technical solutions of the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0019] Figure 1 This is a flowchart illustrating a high-resolution imaging method for sparse aperture radar based on compressed sensing disclosed in this application. Figure 2 This is a schematic diagram illustrating the selection of a target transmitting array element and a target receiving array element as disclosed in this application; Figure 3 This is a schematic diagram of a sparse echo data acquisition method disclosed in this application; Figure 4 This is a schematic diagram of target plane coordinate intersection and positioning disclosed in this application; Figure 5 This is a schematic diagram of the structure of a sparse aperture radar high-resolution imaging system based on compressed sensing disclosed in this application. Detailed Implementation

[0020] To enable those skilled in the art to better understand the present application, the present application will be further described in detail below with reference to the accompanying drawings and specific embodiments. Obviously, the described embodiments are merely some embodiments of the present application, and not all embodiments. Based on the embodiments in this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0021] Figure 1 This is a flowchart illustrating a high-resolution imaging method for sparse aperture radar based on compressed sensing disclosed in this application. The method includes: S101. Based on the target sparsity of the target scene, the spatial distribution of radar elements in the antenna array, and the coherence constraints of the candidate antenna subarrays, select the target transmitting elements and target receiving elements that participate in the operation in the antenna array, so as to select the corresponding operating transmitting channel and operating receiving channel.

[0022] The target scene includes the observation area such as the ground area, vehicle area, or building area that the radar needs to image; target sparsity is used to represent the sparsity of the main scattering points in the target scene. The main scattering points can be understood as the reflection positions where the echo energy is greater than or equal to a preset energy threshold and is retained to participate in imaging; the spatial distribution of radar elements in the antenna array includes the actual installation position of each radar element in the array coordinate system; candidate antenna subarrays include several candidate transmitting elements and candidate receiving elements that can be selected from the complete antenna array; coherence constraints are used to limit the observation results formed by different radar elements from being too similar, so as to avoid confusing the target scattering points at different locations during subsequent imaging.

[0023] like Figure 2 As shown, there are multiple different radar elements 201 in the antenna array. The radar platform sets the upper limit of the channel according to the target sparsity of the target scene, and uses the cross-correlation coefficient corresponding to the candidate antenna subarray as the coherence constraint to filter the candidate transmitting elements and candidate receiving elements, so as to obtain the target transmitting elements labeled as transmitting and the target receiving elements labeled as receiving.

[0024] In this embodiment of the application, S101 specifically includes: S1011. Based on the target sparsity of the target scene, set the upper limit value of the channels that can be opened simultaneously.

[0025] The upper limit of the channel is used to limit the maximum number of working transmit channels and working receive channels that can be opened within the same acquisition period. This upper limit of the channel is not a fixed constant, but is set according to the number of main scattering points in the target scene, the channel capability of the radar platform, and the imaging processing burden.

[0026] In this embodiment, a coarse echo image of the target scene is first acquired through low-resolution pre-scanning. Then, an energy threshold segmentation method is used to identify the main scattering regions. This method retains regions in the echo image with energy values ​​greater than or equal to a preset energy threshold and uses these regions as candidate regions for main scattering points. Next, the radar platform determines the target sparsity based on the estimated number of main scattering points and sets upper channel values ​​according to preset channel configuration rules. When the target sparsity is less than or equal to the preset sparsity threshold, the control unit selects the first channel upper limit value. When the target sparsity is greater than the preset sparsity threshold, the control unit selects a second channel upper limit value that is greater than the first channel upper limit value but does not exceed the radar platform's channel capacity, ensuring that the sparse echo data meets the observation quantity required for subsequent phase residual calculation. The preset energy threshold, preset sparsity threshold, first channel upper limit value, and second channel upper limit value can be preset based on radar platform calibration results, the number of radio frequency channels, and imaging task requirements. S1012. Calculate the spatial pointing characteristics of each radar element in the candidate antenna subarray based on the spatial distribution location, and perform correlation analysis on the spatial pointing characteristics corresponding to different radar elements in the candidate antenna subarray to obtain the cross-correlation coefficient used to characterize the coherence constraint.

[0027] Among them, the spatial pointing feature is used to represent the directional difference formed when a radar element observes the target scene from its spatial distribution position. In other words, different radar elements have different propagation paths and phase changes when facing the same target scattering point due to different installation positions. The cross-correlation number is used to represent the similarity of the spatial pointing features of two radar elements. The larger the cross-correlation number, the more similar the information provided by the two elements.

[0028] In this embodiment, the target scene is first divided into multiple two-dimensional imaging grid nodes, and the propagation path change from each radar element to each grid node is used as the spatial pointing feature. Then, each spatial pointing feature is normalized so that the features of different elements are within a comparable scale. Then, the normalized cross-correlation analysis method is used to calculate the similarity of the spatial pointing features of any two radar elements in the candidate antenna subarray, and the cross-correlation coefficient is obtained. This cross-correlation coefficient is used to characterize the coherence constraint and serves as the basis for subsequent screening of target transmitting elements and target receiving elements.

[0029] S1013. Within the upper limit of the channel, using the minimization of the cross-correlation coefficient as the screening condition, select the target transmitting array element and the target receiving array element with non-uniform physical spacing from the candidate antenna subarray.

[0030] The non-uniform distribution of physical spacing refers to the fact that the adjacent distances between the selected target transmitting array element and the target receiving array element in the antenna array are not exactly the same. The reason for adopting a non-uniform distribution of physical spacing is that array element combinations with completely equal spacing are prone to forming repeated observation relationships in certain directions, while non-uniform spacing can increase the observation difference, which is beneficial to subsequent compressed sensing reconstruction and phase unwrapping.

[0031] In this embodiment, a stepwise greedy selection algorithm can be used to complete the selection: First, calculate the number of grid nodes that can be covered by each candidate transmitting element and candidate receiving element in the candidate antenna subarray for the target scene, and select a group of candidate transmitting elements and candidate receiving elements whose number of covered grid nodes meets the preset coverage requirement as the initial element combination; Second, try to add new elements one by one to the unselected candidate radar elements, and recalculate the maximum cross-correlation number after the addition; If the addition of a candidate radar element makes the maximum cross-correlation number less than the maximum cross-correlation number before the addition, or if the maximum cross-correlation number is less than or equal to the preset cross-correlation threshold while there are at least two different interval values ​​in the adjacent physical distance sequence of the selected element, then the candidate radar element is retained; Then repeat the above process until the number of selected target transmitting elements and target receiving elements reaches the channel upper limit, or the maximum cross-correlation number is less than or equal to the preset cross-correlation threshold; Finally, map the selected target transmitting elements and target receiving elements to the corresponding working transmitting channel and working receiving channel, respectively.

[0032] As an example, when a radar platform images vehicles and fixed facilities along a road, it can first determine whether the number of main scattering points in the target scene is less than or equal to a preset threshold through low-resolution pre-scanning. If the number of main scattering points is less than or equal to the preset threshold, the upper limit of the channel corresponding to that threshold is adopted. Next, the radar platform reads the spatial distribution of each radar element in the antenna array and calculates the spatial pointing characteristics of each candidate radar element facing the road area. If the installation distance between two candidate radar elements is less than a preset element spacing threshold, and the cross-correlation coefficient of their spatial pointing characteristics is greater than a preset cross-correlation threshold, it is determined that the information overlap provided by the two exceeds the screening requirements, and they are not selected simultaneously during screening. If the adjacent distance formed by another candidate radar element and the selected element is different from the existing distance, and its cross-correlation coefficient is less than or equal to a preset cross-correlation threshold, then this candidate radar element can be identified as a target transmitting element or a target receiving element. The above example is only one example of this application. In practical applications, it can also be set according to the radar array size, target scene range, and platform motion conditions. This application does not limit this.

[0033] In another specific implementation, when the number of candidate array elements in the candidate antenna subarray is less than or equal to the preset traversal number threshold, a traversal screening method can also be adopted. That is, the cross-correlation coefficient is calculated for each candidate array element combination that meets the upper limit of the channel, and the combination with the maximum cross-correlation coefficient less than or equal to the preset cross-correlation threshold and the physical spacing is not uniformly distributed is selected. This method is suitable for radar platforms where the number of candidate array element combinations does not exceed the radar platform processing budget.

[0034] This application, through the above process, ensures that the selection of target transmitting array elements and target receiving array elements simultaneously considers the target sparsity of the target scene, the spatial distribution of radar array elements, and the coherence constraints of candidate antenna subarrays. This reduces redundant radio frequency channels, enhances the ability of sparse echo data to distinguish different target scattering points, and provides a more stable data foundation for subsequent high-resolution imaging.

[0035] S102. Control the working transmission channel to sequentially transmit frequency agile signals in a preset frequency jump order at multiple spatial sampling points, and the working receiving channel to receive the echo signals reflected by the target scene, record the sampling position and frequency data corresponding to the echo signals, and obtain sparse echo data.

[0036] Among them, spatial sampling points include discrete positions of the radar platform performing transmission and reception on its motion trajectory; frequency agile signals are used to switch carrier frequencies at different sampling times, so that the same target scene can be observed multiple times in both spatial and frequency dimensions; the preset frequency jump sequence includes the frequency arrangement rules used for each transmission, which are used to indicate the carrier frequency that the working transmission channel should use at each spatial sampling point; sparse echo data includes the correspondence between echo signals, sampling positions and frequency data, which are used to provide input for subsequent phase residual calculation and high-resolution imaging.

[0037] like Figure 3 As shown, the radar platform sequentially passes through multiple spatial sampling points 301 along its motion trajectory. The working transmitting channel transmits frequency-agile signals at each spatial sampling point, and the working receiving channel receives the echo signals formed after reflection from the target scene. The control unit writes the echo signals, along with the corresponding sampling positions, frequency data, and channel information, into the same sparse observation record, thereby forming sparse echo data.

[0038] In this embodiment of the application, S102 specifically includes: S1021. Control the working transmission channel to transmit frequency-agile signals sequentially at the multiple spatial sampling points along the movement trajectory of the radar platform. The carrier frequency of the frequency-agile signal is switched according to the preset frequency jump sequence.

[0039] The radar platform's trajectory includes the continuous path it traverses during imaging; the carrier frequency is the center frequency used when the frequency-agile signal is actually transmitted; the preset frequency transition sequence can be represented by a frequency scheduling table, which records the frequency number and frequency value corresponding to each transmission.

[0040] In this embodiment, the control unit of the radar platform can read multiple spatial sampling points on the motion trajectory and trigger the working transmission channel according to the sampling point sequence. When the transmission conditions are met at each spatial sampling point, the control unit reads the carrier frequency to be used from the frequency scheduling table, then controls the frequency synthesizer to switch to that carrier frequency, and drives the working transmission channel to transmit frequency-agile signals. This process is completed using timing control technology, that is, scheduling the spatial sampling points, transmission time, and carrier frequency in the same time sequence, so that each transmission has a clear spatial location and frequency source.

[0041] S1022. Control the working receiving channel to synchronously receive the echo signal formed after the frequency agile signal is reflected by the target scene.

[0042] Synchronous reception is used to indicate that the working receiving channel opens the receiving window according to the transmission sequence so that the received echo signal can match the corresponding transmission event; the receiving window includes the start receiving time and the end receiving time, which are used to exclude interference signals that are not related to the current transmission event.

[0043] In this embodiment, after the control unit transmits a frequency-agile signal through the working transmission channel, it sets a receiving window according to a preset distance range of the target scene and controls the working receiving channel to collect echo signals within the receiving window. The working receiving channel can amplify, down-convert, and digitally sample the received analog signal to obtain a digital echo signal that can be used for subsequent processing. Since the working receiving channel and the working transmission channel use the same timing trigger information, the received echo signal can be kept consistent with the corresponding spatial sampling point and carrier frequency.

[0044] S1023. Extract the spatial coordinates of the echo signal when it is received as the sampling position, and extract the carrier frequency of the echo signal as frequency data.

[0045] The spatial coordinates include the position coordinates of the radar platform when receiving the echo signal, and may also include the equivalent position of the array elements after attitude compensation; the sampling position is used to indicate the spatial sampling source corresponding to the echo signal; and the frequency data is used to indicate the carrier frequency source corresponding to the echo signal.

[0046] In this embodiment, the spatial coordinates of the received echo signal can be obtained through the positioning unit or inertial navigation unit of the radar platform. If the radar platform experiences attitude changes, the control unit can first use attitude compensation technology to correct the spatial coordinates, and then use the corrected spatial coordinates as the sampling position. At the same time, the control unit reads the carrier frequency used for this transmission from the frequency scheduling table and uses this carrier frequency as the frequency data. Therefore, each echo signal segment can correspond to a specific sampling position and a specific frequency data.

[0047] S1024. Match and bind the echo signal with the corresponding sampling position and frequency data to generate sparse echo data.

[0048] The matching binding is used to write data generated during the same transmission and reception process into the same data record. This data record includes at least the echo signal, sampling position and frequency data, and may further include the transmission channel number, reception channel number and reception time.

[0049] In this embodiment, timestamp matching technology can be used to complete the matching and binding: the control unit generates a transmission timestamp for each frequency-agile signal transmitted and a reception timestamp for the echo signal received by the working receiving channel; then, according to the timing relationship between the transmission timestamp, reception timestamp, and frequency scheduling table, the control unit writes the echo signal and the corresponding sampling position and frequency data into the same sparse observation record. After multiple sparse observation records are combined according to the acquisition order, sparse echo data is formed. Although this sparse echo data does not cover all array elements, all spatial positions, and all frequency points, it retains the key correspondence required for subsequent imaging.

[0050] As an example, the radar platform first moves along a trajectory above the road. When the platform arrives at multiple spatial sampling points in sequence, the control unit controls the working transmission channel to transmit frequency-agile signals according to the frequency scheduling table. For example, the first spatial sampling point uses the first carrier frequency in the frequency scheduling table, and the second spatial sampling point uses another carrier frequency in the frequency scheduling table, so that different spatial sampling points correspond to different frequency observations. Secondly, after each transmission, the working receiving channel collects the echo signal reflected from the target scene according to the receiving window and converts the echo signal into a digital echo signal. Then, the control unit reads the spatial coordinates of the echo signal when it is received as the sampling position and reads the carrier frequency used in this transmission as the frequency data. Finally, the control unit writes the digital echo signal, sampling position, and frequency data into the same sparse observation record, and multiple sparse observation records together constitute sparse echo data. The above example is only one example of this application. In practical applications, settings can be made according to the radar platform speed, target scene range, and frequency resources. This application does not limit these settings.

[0051] In another specific implementation, if the speed of the radar platform is greater than a preset speed threshold, or the displacement between adjacent sampling positions is greater than a preset displacement threshold, the spatial coordinates can be interpolated and corrected first, and then the corrected spatial coordinates can be used as the sampling positions. This method can reduce the position deviation caused by the platform movement and make the sampling positions in the sparse echo data more consistent with the actual reception time.

[0052] Through the above process, this application establishes a clear data correspondence between the working transmit channel, the working receive channel, the sampling position, and the frequency data. This reduces the number of samples while maintaining the traceability of the echo signal, providing a reliable data foundation for subsequent phase unwrapping and high-resolution image generation.

[0053] S103. Based on the first phase breakpoint of the spatial sampling dimension determined by the sampling position and the second phase breakpoint of the frequency jump sequence, a two-dimensional phase constraint is constructed, and the phase periodicity characteristics of the two-dimensional phase constraint are used to unwrap the target scattering point to obtain the first phase trajectory and the second phase trajectory.

[0054] Among them, the first phase breakpoint is used to indicate the phase jump position when the measured phase changes along the spatial sampling dimension; the second phase breakpoint is used to indicate the phase jump position when the measured phase changes along the frequency jump sequence; the two-dimensional phase constraint is used to use the phase changes in both the spatial sampling dimension and the frequency jump sequence to determine the phase continuity relationship; the phase unwrapping is used to restore the measured phase that is restricted to one phase period to a continuously changing phase result.

[0055] In this embodiment of the application, S103 specifically includes: S1031. Extract the echo signal of the sparse echo data within the same distance cell, and retain the wrapped phase of the echo signal to obtain the measurement phase.

[0056] Among them, the distance unit is used to represent the distance position formed after the echo signal is divided according to the propagation distance; the winding phase is used to represent the phase value retained within the preset phase range after the actual phase is periodically folded back; the measured phase is used as the basic data for subsequent calculation of phase residual, phase breakpoint and phase trajectory.

[0057] In this embodiment, the sparse echo data is first subjected to range-directed matched filtering or range compression processing so that the echo signals in the same range resolution unit in the target scene enter the same range unit; then, the complex form echo signals corresponding to each spatial sampling point and each frequency data are extracted from the same range unit; then, the arctangent phase extraction is performed on the complex form echo signals to convert the real and imaginary parts of the complex echo into a wrapped phase within a preset phase range, thereby obtaining the measurement phase limited to one phase period. The measurement phase can be expressed by formula (1): (1) in, Indicates the first The spatial sampling point and the first The measured phase corresponding to each frequency data point; This represents the complex form of the echo signal within the same distance cell; Represents the imaginary part of the echo signal; Represents the real part of the echo signal; This represents the arctangent operation in the four quadrants, used to determine the quadrant in which the phase is located. and All are serial numbers, and the measured phase is expressed in radians.

[0058] S1032. Calculate the phase change of the measured phase between adjacent sampling positions along the spatial sampling dimension to generate a first phase residual, and calculate the phase change of the measured phase between adjacent frequency data along the frequency jump sequence to generate a second phase residual.

[0059] The first phase residual is used to represent the remaining phase change of the measured phase between adjacent sampling positions after deducting the expected phase change; the second phase residual is used to represent the remaining phase change of the measured phase between adjacent frequency data after deducting the expected phase change; the first baseline phase and the second baseline phase are used to represent the expected phase change caused by the reference position, spatial sampling interval and frequency hopping step size.

[0060] In this embodiment of the application, step 1032 may specifically include the following steps: A1. Obtain the spatial sampling interval between adjacent sampling positions and the frequency hopping step size between adjacent frequency data, and construct a first baseline phase and a second baseline phase based on the spatial sampling interval and the frequency hopping step size; A2. Calculate the difference value of the measured phase between adjacent sampling positions along the spatial sampling dimension, and perform phase compensation on the difference value using the first baseline phase to obtain a first phase difference value; A3. Calculate the difference value of the measured phase between adjacent frequency data along the frequency hopping sequence, and perform phase compensation on the difference value using the second baseline phase to obtain a second phase difference value; A4. Perform modulo operation on the first phase difference value and the second phase difference value respectively, and constrain the operation result within the corresponding preset interval to generate a first phase residual and a second phase residual.

[0061] The spatial sampling interval is determined by the coordinate difference between adjacent sampling positions, and the frequency hopping step size is determined by the frequency difference between adjacent frequency data. The construction process of the first baseline phase is as follows: First, using the reference position as a benchmark, calculate the propagation path length under two adjacent spatial sampling points. The propagation path length is the sum of the distance from the target transmitting element to the reference position and the distance from the reference position to the target receiving element. Then, take the difference between the two propagation path lengths, multiply the difference by the current carrier frequency and 2π, and divide by the electromagnetic wave propagation speed to obtain the expected phase change caused by the spatial sampling interval. The construction process of the second baseline phase is as follows: First, read the frequency difference between adjacent frequency data, and then multiply the frequency difference by the propagation path length under the same spatial sampling point. Dividing by the electromagnetic wave propagation speed yields the expected phase change caused by the frequency hopping step size. Therefore, both the spatial sampling interval and the frequency hopping step size are converted into a baseline phase that can be compared with the measured phase.

[0062] The first and second phase residuals can be obtained through adjacent differential calculations, baseline phase compensation, and modulus constraints. Specifically, the measurement phase at the previous sampling position is subtracted from the measurement phase at the subsequent sampling position, and then the first baseline phase is subtracted to obtain the first phase difference. Simultaneously, the measurement phase at the previous frequency data is subtracted from the measurement phase at the subsequent frequency data, and then the second baseline phase is subtracted to obtain the second phase difference. Then, the two phase differences are periodically reversed to ensure they fall within a certain range. to Within the preset interval, the first phase residual and the second phase residual are obtained respectively.

[0063] S1033. The phase jump value generated by the non-uniform sampling interval of the spatial sampling dimension in the first phase residual is determined as the first phase breakpoint, and the phase jump value generated by the frequency jump sequence in the second phase residual is determined as the second phase breakpoint.

[0064] The non-uniform sampling interval is used to indicate that the distances between adjacent sampling positions are not exactly the same. This non-uniformity is reflected in the fact that the distance sequence between adjacent sampling positions is not a fixed value. For example, in a road target scenario, the interval between four consecutive spatial sampling points along the azimuth direction could be 0.40m, 0.65m, and 0.50m respectively, with differences between adjacent intervals. The phase jump value is used to represent the amount of change that occurs when adjacent residuals suddenly cross the phase period boundary. The above intervals are only examples to illustrate the non-uniform sampling interval; the actual interval can be determined based on the radar platform speed, positioning accuracy, and sampling timing.

[0065] In this embodiment, the continuous change of the first phase residual is first checked along the spatial sampling dimension. When the absolute value of the difference between the current spatial sampling interval and the previous spatial sampling interval is greater than a preset interval change threshold, the propagation path will also change accordingly. The first phase residual may jump from one side of the preset interval to the other side. The breakpoint identification unit determines this type of jump position and the corresponding phase jump value as the first phase breakpoint. Secondly, the continuous change of the second phase residual is checked along the frequency hopping sequence. When the absolute value of the difference between the current frequency hopping step size and the previous frequency hopping step size is greater than a preset frequency hopping change threshold, the second phase residual may also jump across cycles. The breakpoint identification unit determines this type of jump position and the corresponding phase jump value as the second phase breakpoint. This processing can be completed using a threshold discrimination method, that is, the position where the residual change exceeds the preset phase change range is taken as the candidate breakpoint, and then confirmed in combination with the change position of the spatial sampling interval or frequency hopping step size.

[0066] S1034. Based on the distribution positions of the first phase breakpoint and the second phase breakpoint in different dimensions, establish a two-dimensional phase constraint for performing phase unwrapping.

[0067] In this embodiment, the first phase breakpoint is first marked in a data grid composed of spatial sampling point numbers and frequency data numbers to indicate an abnormal jump position in the spatial sampling dimension. The second phase breakpoint is then marked in the same data grid to indicate an abnormal jump position in the frequency jump sequence. Next, it is determined whether the first and second phase breakpoints fall within the same data grid region or a data grid region sharing a common boundary. If they correspond within the data grid region corresponding to the same candidate target scattering point, it indicates that the phase jump in that region is supported by both spatial and frequency variations. Finally, this correspondence is written into the phase unwrapping constraint, ensuring that subsequent unwrapping does not rely solely on phase changes in a single direction, but simultaneously satisfies the phase continuity requirements of both the spatial sampling dimension and the frequency jump sequence.

[0068] Specifically, two-dimensional phase constraints can include three types of constraints: the first type is periodic constraints, which limit the phase transitions between the first and second phase breakpoints to an integer number. The first type is the phase period relationship between the two in the same data grid area; the second type is the grid consistency constraint, that is, in the same data grid area, the difference between the phase recovery result obtained along the spatial sampling dimension and the phase recovery result obtained along the frequency jump order does not exceed the preset consistency threshold; the third type is the breakpoint correspondence constraint, that is, the phase step amount used by the first phase breakpoint and the second phase breakpoint pointing to the same candidate target scattering point area should satisfy the same integer period combination.

[0069] S1035. Determine the phase period characteristics based on the phase period between the first phase breakpoint and the second phase breakpoint in the two-dimensional phase constraint.

[0070] Among them, the phase periodicity characteristic is used to indicate that after the measured phase crosses a period boundary, it will increment by an integer number of cycles. The phase transition occurs in a reversed form, a characteristic used to explain that the phase jump is not an arbitrary value, but is related to the phase period.

[0071] In this embodiment, the periodic constraint unit reads the phase transition values ​​corresponding to the first phase breakpoint and the second phase breakpoint, and determines whether these transition values ​​are consistent with the phase transition values. The correspondence between periods. If the first phase breakpoint and the second phase breakpoint are both within the same data grid region, and both satisfy the condition that the phase jump value is a multiple of the nearest integer, then... If the deviation between them is less than or equal to the preset period deviation threshold, the period constraint unit determines the integer period correspondence as the phase period characteristic. This phase period characteristic is then used to limit the unknowns in the untangling equation, so that the untangling equation searches for the phase step only within an integer number of phase periods.

[0072] S1036. By solving the untangling equations corresponding to each phase residual, the phase step lost in the first phase residual and the second phase residual is obtained. The untangling equations are determined based on the phase periodicity characteristics.

[0073] Among them, the phase step is used to represent the integer period compensation amount lost in the first phase residual and the second phase residual due to phase reversal; the untangling equation is used to find the number of phase periods that should be recovered for each residual under two-dimensional phase constraints.

[0074] In this embodiment, the weighted integer least squares algorithm can be used to solve the untangling equation: First, an integer step unknown is set for each first phase residual and second phase residual; second, based on the phase periodicity characteristics, each integer step unknown is converted into a corresponding... Periodic compensation is performed, and the compensated phase residual is denoted as the capped phase residual. Then, with the goal of maintaining consistency between the phase recovery results in the spatial and frequency directions within the same data grid, an error minimization problem is constructed. Finally, this error minimization problem is solved to obtain the lost phase step in the first and second phase residuals. The objective of the untangling equation can be expressed by formula (2): (2) in, Represents the set of integer step unknowns; This means finding the value that minimizes the error within the set of integer step unknowns; Indicates the spatial sampling point number; Indicates the frequency data sequence number; Indicates the first The spatial sampling point and the first The weights corresponding to each frequency data point can be set based on the echo signal strength or signal-to-noise ratio. Indicates the first The spatial sampling point and the first The first phase residual after compensation along the spatial sampling dimension at each frequency data point; Indicates the first The spatial sampling point and the first The second phase residual after compensation along the frequency jump sequence at each frequency data point; Indicates the first The spatial sampling point and the first The second phase residual after compensation along the frequency jump sequence at each frequency data point; Indicates the first The spatial sampling point and the first The first phase residual after spatial sampling dimension compensation at each frequency data point; the expression within the absolute value represents the difference between the accumulated phase along the spatial post-frequency path and the frequency post-spatial path in the same local data grid; the square represents a non-negative error measure of this difference. The smaller the difference, the more consistent the phase recovery results under the two-dimensional phase constraint.

[0075] S1037. The phase step amount is superimposed and compensated to the first phase residual and the second phase residual respectively to reconstruct the continuously monotonically changing first phase trajectory and second phase trajectory.

[0076] The first phase trajectory represents the continuous phase change relationship formed by the target scattering point along the spatial sampling dimension; the second phase trajectory represents the continuous phase change relationship formed by the target scattering point along the frequency jump sequence.

[0077] In this embodiment, the phase step is first added to the first phase residual and the second phase residual respectively to obtain the compensated phase residual; then, the compensated first phase residual is accumulated along the spatial sampling dimension to recover the first phase trajectory; then, the compensated second phase residual is accumulated along the frequency jump sequence to recover the second phase trajectory; finally, the continuity of the first phase trajectory and the second phase trajectory is checked. If the adjacent phase difference value at a local position is greater than a preset continuity threshold, the integer step unknown is readjusted at the corresponding breakpoint position; if the trajectory as a whole maintains continuous monotonic change, the first phase trajectory and the second phase trajectory are output for subsequent target plane coordinate calculation.

[0078] As an example, the radar platform first selects echo signals from the same range cell in the sparse echo data of the road target scene and extracts the measured phase. Next, it calculates the phase change along the spatial sampling dimension and frequency jump sequence, and subtracts the expected phase change caused by the reference position from the first and second baseline phases to obtain the first and second phase residuals. Then, it identifies the phase jump positions caused by the non-uniform sampling interval and frequency jump sequence to obtain the first and second phase breakpoints. Finally, it places the phase breakpoints in the two dimensions into the same data grid for correspondence to determine the phase periodicity characteristics. Finally, it solves for the phase step and compensates it into the first and second phase residuals to obtain the continuously changing first and second phase trajectories.

[0079] In another specific implementation, if the signal-to-noise ratio (SNR) of the echo signal corresponding to a certain sampling position is less than a preset SNR threshold, the weight corresponding to that sampling position can be reduced when solving the unwrapping equation, so that phase unwrapping preferentially utilizes the adjacent regions with an SNR greater than or equal to the preset SNR threshold. This method does not change the basic process of two-dimensional phase constraint, but only changes the degree of influence of different residual data in the solution process.

[0080] This application uses the above process to combine the first phase breakpoint of the spatial sampling dimension and the second phase breakpoint of the frequency jump sequence for phase unwrapping, so that the phase recovery process is simultaneously constrained by spatial and frequency changes, thereby improving the continuity of the first and second phase trajectories and providing a stable basis for subsequent target scattering point positioning.

[0081] S104. Using the intersection information of the first phase trajectory and the second phase trajectory in the two-dimensional imaging plane, calculate the target plane coordinates of the target scattering point, and generate a high-resolution image based on the sparse echo data and the target plane coordinates.

[0082] The two-dimensional imaging plane includes an imaging region composed of azimuth and range directions; the intersection information is used to indicate the position information of the first phase trajectory and the second phase trajectory in the two-dimensional imaging plane that meet the preset intersection conditions or actually intersect; the target plane coordinates are used to indicate the physical position of the target scattering point in the two-dimensional imaging plane.

[0083] like Figure 4 As shown, the two-dimensional imaging plane 401 is used to carry the coordinate grid composed of the azimuth and range directions. The first phase trajectory and the second phase trajectory are projected to form two spatial geometric curves 402. The intersection region 403 of the curves represents the candidate position that simultaneously satisfies the spatial sampling dimension and frequency jump order phase relationship. The coordinate confirmation unit then confirms whether the candidate position corresponds to the target scattering point by comparing the reference phase.

[0084] In this embodiment of the application, S104 specifically includes: S1041. Project the first phase trajectory and the second phase trajectory onto a two-dimensional coordinate grid constructed by the azimuth and range directions, respectively, to generate corresponding spatial geometric curves.

[0085] Among them, the two-dimensional coordinate grid is used to divide the target scene into multiple computable coordinate nodes; the spatial geometric curve is used to represent the set of target positions that the phase trajectory may correspond to in the two-dimensional imaging plane.

[0086] In this embodiment, a two-dimensional coordinate grid consisting of azimuth and range coordinates is first established. Then, for each coordinate node in the two-dimensional coordinate grid, the phase change that the node should produce under the spatial sampling dimension and frequency jump sequence is calculated. Next, the calculated phase changes are matched with the first phase trajectory and the second phase trajectory, respectively. Coordinate nodes that meet the matching conditions are connected to obtain the corresponding spatial geometric curve. Therefore, the first and second phase trajectories are no longer phase data, but are transformed into geometric position cues that can be used for localization in the two-dimensional imaging plane.

[0087] S1042. Calculate the vertical distance between the spatial geometric curves, and take the physical coordinate point corresponding to the intersection position where the vertical distance reaches the minimum value as the initial plane coordinate.

[0088] The vertical distance represents the shortest distance between two spatial geometric curves in the two-dimensional imaging plane; the initial planar coordinates represent the position of the candidate target that has not yet been verified by the reference phase.

[0089] In this embodiment, a grid search algorithm can be used to calculate the vertical distance between spatial geometric curves. Specifically: first, a grid node is selected on the spatial geometric curve corresponding to the first phase trajectory; second, the point with the smallest distance value to the grid node is found on the spatial geometric curve corresponding to the second phase trajectory, and the distance between the two is taken as the vertical distance corresponding to the grid node; then, the above process is repeated for each grid node on the spatial geometric curve corresponding to the first phase trajectory; finally, the position where the vertical distance reaches the minimum value is taken as the intersection position, and the physical coordinate point corresponding to the intersection position is taken as the initial plane coordinate.

[0090] S1043. Calculate the corresponding reference phase based on the initial plane coordinates, the spatial sampling points, and the frequency transition sequence.

[0091] The reference phase is used to represent the theoretical phase that should occur at each spatial sampling point and at each frequency if the initial planar coordinates are indeed the target scattering point.

[0092] In this embodiment, the initial plane coordinates are first read; then, spatial sampling points are read according to the acquisition order in the sparse echo data, and the corresponding frequency data are read according to the preset frequency jump order, so that each spatial sampling point and the frequency data used in the actual transmission form a one-to-one data pair; then, for each data pair, the propagation path length from the target transmitting element to the initial plane coordinates and then to the target receiving element is calculated; finally, the propagation path length and the carrier frequency in the data pair are converted into a reference phase.

[0093] S1044. Compare the difference between the reference phase and the measured phase. If the difference is less than or equal to a preset convergence threshold, then determine that the initial plane coordinates correspond to the target scattering point.

[0094] The difference result is used to represent the magnitude of the phase error between the reference phase and the measured phase; the preset convergence threshold is used to determine whether the initial plane coordinates meet the phase consistency requirements. When the difference result is less than or equal to the preset convergence threshold, it indicates that the error between the reference phase and the measured phase is within the allowable range.

[0095] In this embodiment, the reference phase and the measured phase are first subtracted by a ring phase to obtain the phase error; then, the phase errors corresponding to multiple spatial sampling points and multiple frequency data are weighted and averaged to obtain the difference result; finally, if the difference result is less than or equal to a preset convergence threshold, the target scattering point corresponding to the initial plane coordinates is determined; if the difference result is greater than the preset convergence threshold, other intersection positions are searched, and the difference result can be expressed by formula (3): (3) in, This represents the difference between the initial plane coordinates; ; Indicates the first The spatial sampling point and the first The weights corresponding to each frequency data point; This represents the phase error obtained by subtracting the reference phase and the measured phase using a circular phase calculation. The smaller the value, the closer the reference phase is to the measured phase.

[0096] S1045. The initial plane coordinates are confirmed as the target plane coordinates of the target scattering point.

[0097] In this embodiment of the application, the coordinate confirmation unit writes the verified initial plane coordinates into the target plane coordinate set. If there are multiple target scattering points in the target scene, the coordinate confirmation unit can repeat the processing of S1042 to S1044 for multiple intersection positions and confirm the multiple initial plane coordinates that meet the preset convergence threshold as target plane coordinates respectively. If there is only one valid intersection position, only one target plane coordinate is retained in the target plane coordinate set.

[0098] S1046. Obtain the initial scattering intensity corresponding to each target plane coordinate in the target plane coordinate set.

[0099] The initial scattering intensity is used to represent the estimated echo intensity of the target scattering point before pattern gain compensation.

[0100] In this embodiment, the initial scattering intensity can be obtained by a phase alignment accumulation method: First, the reference phase corresponding to the target plane coordinates is read; second, the echo signals related to the target plane coordinates in the sparse echo data are phase aligned according to the reference phase so that the echo contributions from the same target scattering point can be superimposed; then, the phase-aligned echo signals are weighted and averaged to obtain the initial scattering intensity, which can be expressed by formula (4): (4) in, Indicates the first The initial scattering intensity corresponding to the plane coordinates of each target; Represents the first in the target plane coordinate set The target plane coordinates; Indicates the first The spatial sampling point and the first The echo signal corresponding to each frequency data; This indicates the weight corresponding to the echo signal; Indicates exponentiation; Represents the imaginary unit; Represents the target plane coordinates In the The spatial sampling point and the first The reference phase corresponding to each frequency data point.

[0101] S1047. Based on the spatial distribution of radar elements in the antenna array, calculate the pattern gain at the target plane coordinates, and use the pattern gain to perform gain compensation on the initial scattering intensity to calculate the target scattering intensity.

[0102] In this embodiment of the application, S1047 may specifically include the following steps: B1. Calculate the relative spatial distance between each radar element and each target plane coordinate based on the spatial distribution position of each radar element; B2. Calculate the spatial weighting coefficients projected by each radar element onto each target plane coordinate based on the relative spatial distances; B3. Coherently superimpose the spatial weighting coefficients corresponding to each radar element to calculate the radiation pattern gain at the target plane coordinates; B4. Use the reciprocal of the radiation pattern gain as the gain compensation coefficient, and multiply the gain compensation coefficient by the initial scattering intensity to calculate the target scattering intensity.

[0103] In this context, each target plane coordinate represents every coordinate in the target plane coordinate set; relative spatial distance is used to represent the propagation distance relationship between the radar element and the target plane coordinate; spatial weighting coefficient is used to represent the weight of the radar element's contribution to the echo projected onto the target plane coordinate; target scattering intensity is used to represent the scattering intensity after pattern gain compensation; and pattern gain is used to represent the energy focusing capability formed by the radar element in the antenna array at a certain target plane coordinate.

[0104] Specifically, the spatial distribution positions of each target transmitting element, each target receiving element, and the target plane coordinate set are read. For any target plane coordinate in the target plane coordinate set, the distance from the target transmitting element to that target plane coordinate and the distance from that target plane coordinate to the target receiving element are calculated respectively. The two distances are added together to obtain the relative spatial distance between the radar element channel and the target plane coordinate. If there are multiple target plane coordinates, the above process is repeated for each target plane coordinate to form a set of relative spatial distances between each radar element channel and each target plane coordinate. Then, spatial weighting coefficients are calculated based on the relative spatial distance, element directional weights, and phase alignment relationships. The spatial weighting coefficients are then coherently superimposed to obtain the radiation pattern gain. When using the radiation pattern gain to compensate for the initial scattering intensity, if the radiation pattern gain is zero or less than a preset lower gain limit, the preset lower gain limit is used as the compensation denominator to avoid distortion of the compensation result due to the radiation pattern gain being lower than the preset lower gain limit.

[0105] S1048. Assign the target scattering intensity to the nodes corresponding to the target plane coordinates in the preset two-dimensional grid to generate a high-resolution image.

[0106] The preset two-dimensional grid is used to carry the final imaging result; the high-resolution image includes the scattering intensity distribution corresponding to each grid node.

[0107] In this embodiment, the target plane coordinate set and the target scattering intensity corresponding to each target plane coordinate are first read; then, the node corresponding to each target plane coordinate is found in the preset two-dimensional grid; then, the node is assigned the corresponding target scattering intensity; finally, the grid nodes with unconfirmed target scattering points are assigned background values ​​or zero values, and the scattering intensity distribution is output according to the two-dimensional grid position to generate a high-resolution image.

[0108] As an example, the positional cues obtained from the first and second phase trajectories in the road target scene are first projected onto a two-dimensional coordinate grid composed of azimuth and range directions, forming two spatial geometric curves. Next, the vertical distance between the two spatial geometric curves is calculated, and the intersection position with the smallest vertical distance is used as the initial planar coordinates. Then, a reference phase is calculated by combining spatial sampling points and frequency jump order, and the difference between the reference phase and the measured phase is compared. When the difference is less than or equal to a preset convergence threshold, the target scattering point corresponding to the initial planar coordinates is confirmed. Then, the target planar coordinates are determined and the initial scattering intensity is obtained. Finally, pattern gain compensation is performed based on the spatial distribution of radar array elements, and the target scattering intensity is written into the two-dimensional grid to generate a high-resolution image.

[0109] In another specific implementation, if there are multiple intersection points of the spatial geometric curve with a vertical distance less than or equal to a preset candidate distance threshold, they can be sorted in ascending order of vertical distance, and then reference phase comparison can be performed sequentially. In this way, when there are multiple target scattering points in the target scene, multiple target plane coordinates can be formed, and the corresponding target scattering intensity can be calculated for each.

[0110] This application transforms the intersection information of the first phase trajectory and the second phase trajectory into the target plane coordinates of the target scattering point through the above process, and generates a high-resolution image by combining sparse echo data and pattern gain compensation, thereby improving the stability of target scattering point localization and image intensity representation.

[0111] This application first selects the working transmit and receive channels based on target sparsity, the spatial distribution of radar elements, and the coherence constraints of candidate antenna subarrays through steps S101-S104. Then, sparse echo data is acquired under spatial sampling points and frequency hopping sequences. Subsequently, a two-dimensional phase constraint is constructed using the first and second phase breakpoints, and phase unwrapping is completed. Finally, the target plane coordinates are determined through the intersection information of the first and second phase trajectories, and a high-resolution image is generated by combining this with pattern gain compensation. Therefore, while reducing the channel and sampling burden, the stability of target scattering point localization and image intensity representation in sparse aperture radar imaging can be improved.

[0112] Figure 5 This is a schematic diagram of the structure of a sparse aperture radar high-resolution imaging system based on compressed sensing disclosed in this application, with reference to... Figure 5 The system may include: The selection module 51 is used to select target transmitting elements and target receiving elements participating in the operation in the antenna array based on the target sparsity of the target scene, the spatial distribution position of radar elements in the antenna array, and the coherence constraints of candidate antenna subarrays, so as to select the corresponding working transmitting channel and working receiving channel. The recording module 52 is used to control the working transmission channel to sequentially transmit frequency agile signals in a preset frequency switching order at multiple spatial sampling points, and the working receiving channel to receive the echo signal reflected by the target scene, record the sampling position and frequency data corresponding to the echo signal, and obtain sparse echo data. The construction module 53 is used to construct a two-dimensional phase constraint based on the first phase breakpoint of the spatial sampling dimension determined by the sampling position and the second phase breakpoint of the frequency jump sequence, and to use the phase periodicity characteristics of the two-dimensional phase constraint to perform phase unwrapping on the target scattering point to obtain the first phase trajectory and the second phase trajectory. The generation module 54 is used to calculate the target plane coordinates of the target scattering point using the intersection information of the first phase trajectory and the second phase trajectory in the two-dimensional imaging plane, and generate a high-resolution image based on the sparse echo data and the target plane coordinates.

[0113] The high-resolution imaging system based on compressed sensing sparse aperture radar in this application is used to implement the aforementioned high-resolution imaging method based on compressed sensing sparse aperture radar. Therefore, the specific implementation of the high-resolution imaging system based on compressed sensing sparse aperture radar can be found in the embodiment section of the high-resolution imaging method based on compressed sensing sparse aperture radar mentioned above. The specific implementation can be referred to the description of the corresponding embodiments, which will not be repeated here.

[0114] Those skilled in the art will further recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of both. To clearly illustrate the interchangeability of hardware and software, the components and steps of the various examples have been generally described in terms of functionality in the foregoing description. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0115] The above provides a detailed description of a high-resolution imaging method and system for sparse aperture radar based on compressed sensing, as provided in this application. Specific examples have been used to illustrate the principles and implementation methods of this application. The descriptions of the embodiments above are merely for the purpose of helping to understand the method and its core ideas. It should be noted that those skilled in the art can make various improvements and modifications to this application without departing from its principles, and these improvements and modifications also fall within the protection scope of this application.

Claims

1. A method for high-resolution imaging of sparse-aperture radar based on compressed sensing, applied to a radar platform, characterized in that, include: Based on the target sparsity of the target scene, the spatial distribution of radar elements in the antenna array, and the coherence constraints of the candidate antenna subarrays, target transmitting elements and target receiving elements participating in the operation are selected in the antenna array to select the corresponding operating transmitting channel and operating receiving channel. The working transmitting channel is controlled to sequentially transmit frequency agile signals in a preset frequency switching order at multiple spatial sampling points, and the working receiving channel receives the echo signals reflected from the target scene, records the sampling position and frequency data corresponding to the echo signals, and obtains sparse echo data. Based on the first phase breakpoint of the spatial sampling dimension determined by the sampling position and the second phase breakpoint of the frequency jump sequence, a two-dimensional phase constraint is constructed, and the phase periodicity characteristics of the two-dimensional phase constraint are used to unwrap the target scattering point to obtain the first phase trajectory and the second phase trajectory. Using the intersection information of the first phase trajectory and the second phase trajectory in the two-dimensional imaging plane, the target plane coordinates of the target scattering point are calculated, and a high-resolution image is generated based on the sparse echo data and the target plane coordinates.

2. The method according to claim 1, characterized in that, The construction of a two-dimensional phase constraint based on the first phase breakpoint of the spatial sampling dimension determined by the sampling position and the second phase breakpoint of the frequency jump sequence includes: The echo signal of the sparse echo data within the same distance cell is extracted, and the wrapped phase of the echo signal is retained to obtain the measurement phase; The phase change of the measured phase between adjacent sampling positions is calculated along the spatial sampling dimension to generate a first phase residual, and the phase change of the measured phase between adjacent frequency data is calculated along the frequency jump sequence to generate a second phase residual. The phase jump value generated by the non-uniform sampling interval of the spatial sampling dimension in the first phase residual is determined as the first phase breakpoint, and the phase jump value generated by the frequency jump sequence in the second phase residual is determined as the second phase breakpoint. Based on the distribution positions of the first phase breakpoint and the second phase breakpoint in different dimensions, a two-dimensional phase constraint is established for performing phase unwrapping.

3. The method according to claim 2, characterized in that, The step of calculating the phase change of the measured phase between adjacent sampling positions along the spatial sampling dimension to generate a first phase residual, and calculating the phase change of the measured phase between adjacent frequency data along the frequency jump sequence to generate a second phase residual, includes: The spatial sampling interval between adjacent sampling positions and the frequency hopping step size between adjacent frequency data are obtained respectively. A first baseline phase and a second baseline phase are constructed based on the spatial sampling interval and the frequency hopping step size. The difference value of the measured phase between adjacent sampling positions is calculated along the spatial sampling dimension, and the first baseline phase is used to perform phase compensation on the difference value to obtain the first phase difference value; The difference value of the measured phase between adjacent frequency data is calculated along the frequency jump sequence, and the second baseline phase is used to perform phase compensation on the difference value to obtain the second phase difference value; Modulo operations are performed on the first phase difference and the second phase difference respectively, and the operation results are constrained within the corresponding preset intervals to generate the first phase residual and the second phase residual.

4. The method according to claim 1, characterized in that, The step of using the phase periodicity characteristic of the two-dimensional phase constraint to perform phase unwrapping on the target scattering point to obtain the first phase trajectory and the second phase trajectory includes: The phase period characteristics are determined based on the phase period between the first phase breakpoint and the second phase breakpoint in the two-dimensional phase constraint. By solving the unwrapping equations corresponding to each phase residual, the phase step lost in the first and second phase residuals is obtained. The unwrapping equations are determined based on the phase periodicity characteristics. The phase step is superimposed and compensated to the first phase residual and the second phase residual respectively, to reconstruct the continuously monotonically changing first phase trajectory and second phase trajectory.

5. The method according to claim 1, characterized in that, The step of calculating the target plane coordinates of the target scattering point using the intersection information of the first phase trajectory and the second phase trajectory in the two-dimensional imaging plane includes: The first phase trajectory and the second phase trajectory are projected onto a two-dimensional coordinate grid constructed by the azimuth and range directions, respectively, to generate corresponding spatial geometric curves; Calculate the vertical distance between the spatial geometric curves, and take the physical coordinate point corresponding to the intersection position where the vertical distance reaches the minimum value as the initial plane coordinates; Based on the initial plane coordinates, combined with the spatial sampling points and the frequency transition sequence, the corresponding reference phase is calculated; The reference phase is compared with the measured phase. If the difference is less than or equal to a preset convergence threshold, the initial plane coordinates are determined to correspond to the target scattering point. The initial plane coordinates are confirmed as the target plane coordinates of the target scattering point.

6. The method according to claim 1, characterized in that, The generation of a high-resolution image based on the sparse echo data and the target plane coordinates includes: Obtain the initial scattering intensity corresponding to each target plane coordinate in the target plane coordinate set; Based on the spatial distribution of radar elements in the antenna array, the pattern gain at the target plane coordinates is calculated, and the initial scattering intensity is compensated using the pattern gain to calculate the target scattering intensity. In a preset two-dimensional grid, the nodes corresponding to the target plane coordinates are assigned the target scattering intensity to generate a high-resolution image.

7. The method according to claim 6, characterized in that, The step of calculating the radiation pattern gain at the target plane coordinates based on the spatial distribution of radar elements in the antenna array, and using the radiation pattern gain to perform gain compensation on the initial scattering intensity to calculate the target scattering intensity includes: Calculate the relative spatial distance between each radar element and the plane coordinates of each target based on the spatial distribution of each radar element. Based on the relative spatial distance, calculate the spatial weighting coefficients of the projection of each radar element onto the plane coordinates of each target; The spatial weighting coefficients corresponding to the radar array elements are coherently superimposed to calculate the radiation pattern gain at the target plane coordinates. The reciprocal of the pattern gain is used as the gain compensation coefficient, and the gain compensation coefficient is multiplied by the initial scattering intensity to calculate the target scattering intensity.

8. The method according to claim 1, characterized in that, The control of the working transmitting channel to sequentially transmit frequency-agile signals in a preset frequency-hopping order at multiple spatial sampling points, and the receiving channel to receive the echo signal reflected from the target scene, includes: The working transmission channel is controlled to move along the trajectory of the radar platform and transmit frequency-agile signals sequentially at the multiple spatial sampling points. The carrier frequency of the frequency-agile signal is switched according to the preset frequency jump sequence. The working receiving channel is controlled to synchronously receive the echo signal formed after the frequency agile signal is reflected by the target scene; The spatial coordinates of the echo signal when it is received are extracted as the sampling position, and the carrier frequency of the echo signal is extracted as frequency data. The echo signal is matched and bound with the corresponding sampling position and frequency data to generate sparse echo data.

9. The method according to claim 1, characterized in that, The step of selecting target transmitting and receiving elements from the antenna array based on target sparsity, spatial distribution of radar elements in the antenna array, and coherence constraints of candidate antenna subarrays includes: Based on the target sparsity of the target scenario, set the upper limit value of the channels that can be opened simultaneously. The spatial pointing characteristics of each radar element in the candidate antenna subarray are calculated based on the spatial distribution location, and the correlation analysis is performed on the spatial pointing characteristics corresponding to different radar elements in the candidate antenna subarray to obtain the cross-correlation coefficient used to characterize the coherence constraint. Within the upper limit of the channel, the target transmitting element and the target receiving element with non-uniform physical spacing are selected from the candidate antenna subarrays by using the minimization of the cross-correlation coefficient as the screening condition.

10. A high-resolution imaging system for sparse aperture radar based on compressed sensing, characterized in that, include: The selection module is used to select target transmitting elements and target receiving elements in the antenna array based on the target sparsity of the target scene, the spatial distribution of radar elements in the antenna array, and the coherence constraints of candidate antenna subarrays, so as to select the corresponding working transmitting channel and working receiving channel. The recording module is used to control the working transmission channel to sequentially transmit frequency agile signals in a preset frequency switching order at multiple spatial sampling points, and the working receiving channel to receive the echo signals reflected by the target scene, record the sampling position and frequency data corresponding to the echo signals, and obtain sparse echo data. The construction module is used to construct a two-dimensional phase constraint based on the first phase breakpoint of the spatial sampling dimension determined by the sampling position and the second phase breakpoint of the frequency jump sequence, and to use the phase periodicity characteristics of the two-dimensional phase constraint to perform phase unwrapping on the target scattering point to obtain the first phase trajectory and the second phase trajectory. The generation module is used to calculate the target plane coordinates of the target scattering point using the intersection information of the first phase trajectory and the second phase trajectory in the two-dimensional imaging plane, and generate a high-resolution image based on the sparse echo data and the target plane coordinates.

Images