A three-dimensional target electromagnetic property perception method based on a reconfigurable intelligent surface

Abstract

The application relates to the technical field of electromagnetic imaging, in particular to a three-dimensional target electromagnetic property perception method based on a reconfigurable intelligent surface, which comprises the following steps: configuring parameters of an electromagnetic imaging system, generating RIS unit scattering coefficients based on mutual coherence optimization by adopting an MCBB algorithm; configuring the scattering coefficients on the RIS, scattering and reconstructing electromagnetic waves generated by a transmitting antenna; linearly correlating contrast parameters of a measurement target with received scattering field data by adopting a linear measurement model; and finally, inverting the target electromagnetic property by utilizing mt_cs. The application realizes high-precision inversion of the target electromagnetic property in the RIS-assisted imaging framework by optimizing and designing the super-surface reflection coefficient and constructing a measurement matrix with multi-dimensional low cross-correlation characteristics.

Description

Technical Field

[0001] This invention relates to the field of electromagnetic imaging technology, specifically to a method for sensing the electromagnetic properties of a three-dimensional target based on a reconfigurable smart surface. Background Technology

[0002] In recent years, Reconfigurable Smart Surfaces (RIS) have become an important hardware platform for building "programmable beam illumination" due to their ultra-thin shape, low power consumption, and flexible modulation capabilities of electromagnetic beams. Unlike traditional electromagnetic imaging, which relies on antenna scanning and array synthetic aperture as incident electromagnetic sources for sensing electromagnetic properties, RIS-assisted imaging emphasizes encoding and measuring targets through reconfigurable electromagnetic field patterns, eliminating the need for additional electromagnetic source settings. In other words, RIS-assisted imaging can serve as a novel imaging system that can replace static electromagnetic sources in scenarios where they are fixed.

[0003] RIS-assisted electromagnetic imaging systems control the reflection coefficients of RIS cells to modulate the distribution of the incident electromagnetic field, generating spatially diverse illumination patterns. This, combined with computational imaging algorithms, enables efficient reconstruction of the target's electromagnetic properties. Traditional methods typically use random Gaussian matrices to encode RIS cells, leveraging the low correlation between different measurements to effectively perceive the target's electromagnetic properties. Some researchers have introduced ghost imaging into RIS-assisted electromagnetic imaging, pre-constructing strictly orthogonal ideal measurement patterns and further inverting the corresponding RIS cell control states to approximate an ideal electromagnetic mask on the target region, achieving inversion reconstruction. However, limited by the physical implementation capabilities of RIS devices and electromagnetic propagation characteristics, both existing imaging strategies have significant limitations in practical applications.

[0004] I. Although the random Gaussian matrix-based coding method (Rand) can reduce the cross-correlation of reflectance coefficients between different measurements at the RIS unit level, this low correlation is often difficult to maintain effectively after forward propagation model mapping, resulting in strong correlation of illumination patterns in the target domain, which limits the independence of measurement information and imaging reconstruction performance.

[0005] II. Although the orthogonal measurement mode design method (GI) based on the ghost imaging concept can theoretically improve reconstruction stability by constructing strictly orthogonal ideal measurement modes, it is difficult to accurately synthesize such ideal modes due to the limitations of the physical implementation capabilities of actual RIS devices. At the same time, the number of measurements required usually far exceeds the number of target voxels, thus resulting in low data acquisition efficiency.

[0006] To this end, we developed a computational imaging method that aims to reduce the cross-correlation of measurement modes as the design goal of RIS, while taking into account the requirements of low measurement frequency and high robustness. Summary of the Invention

[0007] The purpose of this invention is to provide a three-dimensional target electromagnetic property sensing method based on reconfigurable smart surfaces, so as to solve the problems of high cross-correlation of measurement modes, difficulty in hardware implementation, and low acquisition efficiency in traditional methods.

[0008] To solve the above-mentioned technical problems, the present invention provides the following technical solution:

[0009] A method for sensing the electromagnetic properties of a three-dimensional target based on a reconfigurable smart surface includes the following steps:

[0010] Step 1: Construct an electromagnetic sensing system consisting of a transmitting antenna, a reconfigurable smart surface (RIS), a target area, and a receiving antenna, and configure the system parameters, including the operating frequency, wavelength, RIS unit parameters, transmitting antenna position, receiving antenna position, and discrete grid parameters of the target area.

[0011] Step 2: Generate an incident electric field through the transmitting antenna. With the incident magnetic field The MCBB algorithm is used to generate a scattering coefficient matrix based on mutual coherence optimization. To reduce the correlation of the perception matrix and improve the stability of subsequent reconstruction;

[0012] Step 3: Scattering coefficient matrix A reconfigurable smart surface (RIS) is configured, and the electromagnetic waves generated by the transmitting antenna are reconstructed and scattered by the RIS to generate a RIS scattering field. The method of moments is used to solve for the distribution of the RIS scattering field on the discrete grid of the target region to obtain the discretized RIS scattering field of the target region. ;

[0013] Step 4: The RIS scattered field is scattered by objects in the target area, and the receiving antenna scattered field is obtained at the receiving antenna. The received scattered field The electromagnetic properties of the target and the discretized RIS scattering field of the target region The propagation relationship from the target area to the receiving antenna is also determined by this;

[0014] Step 5: Based on the scattering coefficient matrix Discretized lighting field for target area and receiving scattered fields A sensing equation is constructed between the electromagnetic properties of the target and the received scattered field, and a multi-target Bayesian compressed sensing reconstruction algorithm based on compressed sensing theory is used for reconstruction.

[0015] Preferably, in step 2, the incident electric field With the incident magnetic field The specific formula is:

[0016] ;

[0017] in, For free space wave impedance, The wavenumber in the background medium. This is the elevation angle of the transmitting antenna.

[0018] Preferably, in step 2, the MCBB algorithm is used to generate the scattering coefficient matrix based on mutual coherence optimization. The process includes the following steps:

[0019] Step 2.1: The RIS scattered field propagates to the target region with a linear mapping relationship. The propagation matrix is ​​obtained by discretizing this process using the method of moments. The process is described as follows: The propagation matrix With RIS scattering coefficient matrix Transpose and assign, that is , ;

[0020] Step 2.2, for the matrix Perform singular value decomposition to obtain ;

[0021] Step 2.3, based on the matrix Construct the target correlation matrix The formula is ;

[0022] Step 2.4, for Problem, definition ;

[0023] Step 2.5: Convert the matrix Projected to Constructing an equivalent optimization matrix in Zhang Cheng's eigenspace. ;

[0024] Step 2.6: Optimize the equivalent matrix Perform eigenvalue decomposition to obtain The eigenvalues ​​are arranged in descending order of their real parts as follows: ;

[0025] Step 2.7: Determine the number of valid non-negative eigenvalues. ;

[0026] Step 2.8: Based on the number of measurements Number of valid eigenvalues Based on the size relationship, construct an auxiliary matrix:

[0027] ;

[0028] Step 2.9, according to the definition Reverse reasoning ,in For RIS power.

[0029] Preferably, step 3 specifically includes:

[0030] Approximating RIS as an infinite PEC, we remove the PEC according to the physical equivalence theorem and introduce an equivalent current density on the surface. Using the PO approximation, we get:

[0031] ;

[0032] in, ;

[0033] The formulas for the magnetic vector potential under the Lorentz gauge and the RIS scattered electric field are as follows:

[0034] ;

[0035] in, Representing angular frequency, magnetic permeability, and dielectric constant of the medium, respectively, the scattered field vector of the target region... Discretize and distinguish directions as follows:

[0036] ;

[0037] Among them, superscript They represent Axial direction, and They represent the first The RIS scattering unit and the first There are sampling units, among which ; They represent the first The RIS scattering unit and the first The x-axis and y-axis coordinates of each sampling unit Indicates the first The RIS scattering unit and the first The distance between each sampling unit Indicates the first The second measurement of RIS The reflection coefficient of a small piece;

[0038] After discretization, the three directions of the RIS scattering field are further combined into the same matrix, resulting in: ;where the matrix yes The combined discretized integral kernel of each component is represented as follows:

[0039] .

[0040] Preferably, step 4 specifically includes:

[0041] Assuming the receiving antenna scattering field To accept only The target scattered electric field component along the axis, then the electric field component Represented as: ;

[0042] The three-dimensional free-space Green's function tensor is:

[0043] , This represents the position vector of the receiving point. The position vector representing the center of the target discrete element. Indicates that it is located at along The unit source in the direction produced at the place Directional electric field components; Indicates calculation The Green's function operator for the axially scattered target field, where Let the volume of the target discrete element be denoted as . The total number of units in the three directions after discretizing the target region;

[0044] Represents the contrast vector of the target sampling point Diagonalization, contrast is expressed as: ;in, The relative permittivity, The dielectric constant in vacuum is . For electrical conductivity, ω is the angular frequency of the electromagnetic wave;

[0045] For electric field components Linearization:

[0046] ;

[0047] ;

[0048] Among them, the contrast vector of the target sampling point , This represents the Khatri-Rao product.

[0049] Preferably, step 5 specifically includes:

[0050] Based on the contrast vector of the target sampling points The perceptual equation is expressed as: ;

[0051] in, Given a set of additive Gaussian noise vectors with zero mean and variance varying with the signal-to-noise ratio, As a sparse vector, the multi-objective Bayesian compressed sensing reconstruction algorithm mt-CS based on compressed sensing theory is used to solve it. Reconstructing using norm problems: .

[0052] Preferably, a three-dimensional target electromagnetic property sensing system includes an antenna transmitter, a programmable metasurface, an antenna receiver, and a signal processing unit; the programmable metasurface is used for phase and amplitude modulation of the transmitted wave; the antenna receiver is used for receiving the scattered field; and the signal processing unit is used to execute the MCBB algorithm to design the reflection coefficient of the RIS cell, construct the sensing matrix, and solve for the electromagnetic parameters. This system achieves high-precision sensing and imaging of the electromagnetic properties of three-dimensional targets.

[0053] Preferably, the scattering coefficient matrix is ​​constructed using the MCBB algorithm; the scattering coefficient matrix is ​​configured on a reconfigurable smart surface to reconstruct the incident field phase and amplitude modulation; the receiving antenna is fixed at the position. Used to receive the scattered field and perform compressed sensing MT-CS algorithm inversion.

[0054] Compared with the prior art, the beneficial effects achieved by the present invention are:

[0055] This invention constructs an end-to-end electromagnetic propagation model of “RIS-target region-receiver”, which integrates the incident field propagation operator from RIS to the target region, the scattering Green’s function from the target to the receiver, and the electromagnetic coupling relationship between target voxels into the perception framework. This makes the physical modeling of the target scattering process more complete and can more accurately characterize the distribution characteristics of the target’s electromagnetic contrast in space.

[0056] Compared to beam design methods based on simple random excitation or ghost imaging, this invention combines the MCBB algorithm to optimize the configuration of the RIS scattering matrix, thereby improving the effective illumination capability of the target area under the condition of limited total power, enhancing the information independence and effectiveness of the measurement matrix, and thus improving the quality of target scattering information acquisition and parameter inversion accuracy.

[0057] This invention employs a compressed sensing-based MT-CS reconstruction algorithm to invert the complex electromagnetic contrast of a target. It fully utilizes the sparse distribution characteristics of the target in three-dimensional voxel space to achieve high-precision recovery of the target's dielectric properties. Furthermore, this invention requires only a few receiving channels to acquire target scattering information, reducing system hardware complexity and application costs. It can also output a three-dimensional visualized distribution of the target's electromagnetic contrast, combining high imaging accuracy, strong noise resistance, and good engineering implementation capabilities. Therefore, this invention has high application value in fields such as security imaging, concealed target detection, medium parameter identification, electromagnetic environment sensing, and integrated intelligent communication sensing. Attached Figure Description

[0058] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used in conjunction with embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings:

[0059] Figure 1 A general flowchart of the method for this invention is provided;

[0060] Figure 2 The overall pseudocode for the method of this invention is provided below;

[0061] Figure 3 A schematic diagram of a scenario for providing the method of this invention;

[0062] Figure 4 This invention provides the basic parameter configuration for the method.

[0063] Figure 5 To achieve the sampling rate Comparison of reconstruction results of the method of the present invention and other algorithms on the "inverted T" test target under signal-to-noise ratio conditions of 10dB, 15dB and 20dB;

[0064] Figure 6 To achieve the sampling rate Under the given conditions, the relationship between the reconstruction error of the method of this invention and other algorithms for the "inverted T" test target and the signal-to-noise ratio is shown in the graph.

[0065] Figure 7 To compare the reconstruction error of the proposed method with other algorithms for the "inverted T" test target with the sampling rate under a signal-to-noise ratio of 20dB. A graph showing the changing relationship;

[0066] Figure 8 To achieve the sampling rate Comparison of reconstruction results of the method of the present invention and other algorithms on the three-layer "ABC" test target under signal-to-noise ratio conditions of 10dB, 15dB and 20dB;

[0067] Figure 9 To achieve the sampling rate Under the given conditions, the relationship between the reconstruction error of the method of this invention and other algorithms for the three-layer "ABC" test target and the signal-to-noise ratio is shown in the graph.

[0068] Figure 10 To compare the reconstruction error of the proposed method with other algorithms for a three-layer "ABC" test target with the sampling rate under a signal-to-noise ratio of 20dB. A graph showing the changing relationship;

[0069] Figure 11 To achieve the sampling rate Comparison of the reconstruction effects of the method of the present invention and other algorithms on the "UAV" test target under signal-to-noise ratio conditions of 10dB, 15dB and 20dB;

[0070] Figure 12 To achieve the sampling rate Under the given conditions, the relationship between the reconstruction error of the "UAV" test target by the method of this invention and other algorithms as a function of signal-to-noise ratio is shown in the graph.

[0071] Figure 13 To compare the reconstruction error of the proposed method and other algorithms for the "drone" test target with the sampling rate under a signal-to-noise ratio of 20dB. A graph showing the changing relationship. Detailed Implementation

[0072] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0073] Please see Figures 1-13 The present invention provides the following technical solution:

[0074] Example 1: An electromagnetic property sensing method based on cross-coherence optimized beams with RIS assistance, such as... Figure 1 The process shown includes the following steps:

[0075] Step 1: Construct an electromagnetic sensing system consisting of a transmitting antenna, a reconfigurable smart surface (RIS), a target area, and a receiving antenna, and configure the system parameters, including the operating frequency, wavelength, RIS unit parameters, transmitting antenna position, receiving antenna position, and discrete grid parameters of the target area.

[0076] Step 2: Generate an incident electric field through the transmitting antenna. With the incident magnetic field The MCBB algorithm is used to generate a scattering coefficient matrix based on mutual coherence optimization. To reduce the correlation of the perception matrix and improve the stability of subsequent reconstruction;

[0077] Preferably, in step 2, the incident electric field With the incident magnetic field The specific formula is:

[0078] ;

[0079] in, For free space wave impedance, The wavenumber in the background medium. This is the elevation angle of the transmitting antenna.

[0080] The physical scene diagram of this embodiment is as follows: Figure 3 As shown, the specific general parameter configuration is as follows: Figure 4 The specific data specific to this embodiment are as follows:

[0081] The imaging area is divided into The discrete units, the RIS unit is divided into: Discrete units.

[0082] Step 3: Scattering coefficient matrix A reconfigurable smart surface (RIS) is configured, and the electromagnetic waves generated by the transmitting antenna are reconstructed and scattered by the RIS to generate a RIS scattering field. The method of moments is used to solve for the distribution of the RIS scattering field on the discrete grid of the target region to obtain the discretized RIS scattering field of the target region. ;

[0083] Preferably, step 3 specifically includes:

[0084] Approximating RIS as an infinite PEC, we remove the PEC according to the physical equivalence theorem and introduce an equivalent current density on the surface. Using the PO approximation, we get:

[0085] ;

[0086] in, ;

[0087] The formulas for the magnetic vector potential under the Lorentz gauge and the RIS scattered electric field are as follows:

[0088] ;

[0089] in, Representing angular frequency, magnetic permeability, and dielectric constant of the medium, respectively, the scattered field vector of the target region... Discretize and distinguish directions as follows:

[0090] ;

[0091] Among them, superscript They represent Axial direction, and They represent the first The RIS scattering unit and the first There are sampling units, among which ; They represent the first The RIS scattering unit and the first The x-axis and y-axis coordinates of each sampling unit Indicates the first The RIS scattering unit and the first The distance between each sampling unit Indicates the first The second measurement of RIS The reflection coefficient of a small piece;

[0092] After discretization, the three directions of the RIS scattering field are further combined into the same matrix, resulting in: ;where the matrix yes The combined discretized integral kernel of each component is represented as follows:

[0093] .

[0094] Step 4: The RIS scattered field is scattered by objects in the target area, and the receiving antenna scattered field is obtained at the receiving antenna. The received scattered field The electromagnetic properties of the target and the discretized RIS scattering field of the target region The propagation relationship from the target area to the receiving antenna is also determined by this;

[0095] Preferably, step 4 specifically includes:

[0096] Assuming the receiving antenna scattering field To accept only The target scattered electric field component along the axis, then the electric field component Represented as: ;

[0097] The three-dimensional free-space Green's function tensor is:

[0098] , This represents the position vector of the receiving point. The position vector representing the center of the target discrete element. Indicates that it is located at along The unit source in the direction produced at the place Directional electric field components; Indicates calculation The Green's function operator for the axially scattered target field, where Let the volume of the target discrete element be denoted as . The total number of units in the three directions after discretizing the target region;

[0099] Represents the contrast vector of the target sampling point Diagonalization, contrast is expressed as: ;in, The relative permittivity, The dielectric constant in vacuum is . For electrical conductivity, ω is the angular frequency of the electromagnetic wave;

[0100] For electric field components Linearization:

[0101] ;

[0102] ;

[0103] Among them, the contrast vector of the target sampling point , This represents the Khatri-Rao product.

[0104] Step 5: Based on the scattering coefficient matrix Discretized lighting field for target area and receiving scattered fields A sensing equation is constructed between the electromagnetic properties of the target and the received scattered field, and a multi-target Bayesian compressed sensing reconstruction algorithm based on compressed sensing theory is used for reconstruction.

[0105] Preferably, step 5 specifically includes:

[0106] Based on the contrast vector of the target sampling points The perceptual equation is expressed as: ;

[0107] in, Given a set of additive Gaussian noise vectors with zero mean and variance varying with the signal-to-noise ratio, As a sparse vector, the multi-objective Bayesian compressed sensing reconstruction algorithm mt-CS based on compressed sensing theory is used to solve it. Reconstructing using norm problems: .

[0108] Preferably, the MCBB algorithm process used in this embodiment includes the following steps:

[0109] Step 2.1: The RIS scattered field propagates to the target region with a linear mapping relationship. The propagation matrix is ​​obtained by discretizing this process using the method of moments. The process is described as follows: The propagation matrix With RIS scattering coefficient matrix Transpose and assign, that is , ;

[0110] Step 2.2, for the matrix Perform singular value decomposition to obtain ;

[0111] Step 2.3, based on the matrix Construct the target correlation matrix The formula is ;

[0112] Step 2.4, for Problem, definition ;

[0113] Step 2.5: Convert the matrix Projected to Constructing an equivalent optimization matrix in Zhang Cheng's eigenspace. ;

[0114] Step 2.6: Optimize the equivalent matrix Perform eigenvalue decomposition to obtain The eigenvalues ​​are arranged in descending order of their real parts as follows: ;

[0115] Step 2.7: Determine the number of valid non-negative eigenvalues. ;

[0116] Step 2.8: Based on the number of measurements Number of valid eigenvalues Based on the size relationship, construct an auxiliary matrix:

[0117] ;

[0118] Step 2.9, according to the definition Reverse reasoning ,in For RIS power.

[0119] The result of reconstructing the "inverted T" target using this embodiment is as follows: Figure 5 As shown in the figure, the actual values ​​of the target dielectric parameter profile are compared with the reconstruction results of the method in this embodiment and the comparison method under signal-to-noise ratio conditions of 10dB, 15dB, and 20dB. The target contrast function is set in the figure. .

[0120] Based on this "inverted T" objective, at the sampling rate Under the given conditions, the relationship between the reconstruction error of the method of this invention and other algorithms as a function of signal-to-noise ratio is shown in the following graph. Figure 6 As shown.

[0121] Based on this "inverted T" objective, under a signal-to-noise ratio of 20dB, the reconstruction error of the method of this invention compared to other algorithms decreases with the sampling rate. The relationship curve of change is as follows Figure 7 As shown.

[0122] Example 2: Reconstruction test of three-layer "ABC" targets within a three-dimensional imaging area.

[0123] In this embodiment, the algorithm flow and general parameter configuration are consistent with those of Embodiment 1. The specific data unique to this embodiment are as follows:

[0124] The imaging area is divided into The discrete units, the RIS unit is divided into: Discrete units.

[0125] The results of reconstructing the three-layer "ABC" target are as follows Figure 8 As shown in the figure, the actual values ​​of the target dielectric parameter profile are compared with the reconstruction results of the method and the comparison method in this embodiment under signal-to-noise ratio conditions of 10dB, 15dB, and 20dB. The target contrast function is set in the figure. ;

[0126] Based on these three-layered "ABC" objectives, at the sampling rate Under the given conditions, the relationship between the reconstruction error of the method of this invention and other algorithms as a function of signal-to-noise ratio is shown in the following graph. Figure 9 As shown;

[0127] Based on this three-layer "ABC" target, under a signal-to-noise ratio of 20dB, the reconstruction error of the method of this invention compared with other algorithms decreases with the sampling rate. The relationship curve of change is as follows Figure 10 As shown;

[0128] Example 3: Reconstruction test of "UAV" target within a three-dimensional imaging area.

[0129] In this embodiment, the algorithm flow and general parameter configuration are consistent with those of Embodiment 1. The specific data unique to this embodiment are as follows:

[0130] The imaging area is divided into The discrete units, the RIS unit is divided into: Discrete units.

[0131] The result of reconstructing the "drone" target is as follows Figure 11 As shown in the figure, the actual values ​​of the target dielectric parameter profile are compared with the reconstruction results of the method and the comparison method in this embodiment under signal-to-noise ratio conditions of 10dB, 15dB, and 20dB. The target contrast function is set in the figure. ;

[0132] Based on this "drone" target, at the sampling rate Under the given conditions, the relationship between the reconstruction error of the method of this invention and other algorithms as a function of signal-to-noise ratio is shown in the following graph. Figure 12 As shown;

[0133] Based on the "drone" target, under a signal-to-noise ratio of 20dB, the reconstruction error of the method of this invention compared with other algorithms varies with the sampling rate. The relationship curve of change is as follows Figure 13 As shown.

[0134] Example 4: An electromagnetic property sensing system based on cross-coherence optimized beamforming with RIS assistance, comprising an antenna transmitter, a programmable metasurface, an antenna receiver, and a signal processing unit. The programmable metasurface is used for phase and amplitude modulation of the transmitted wave; the antenna receiver is used to receive the scattered field; and the signal processing unit is used to execute the MCBB algorithm to design the reflection coefficients of the RIS unit, construct the sensing matrix, and solve for electromagnetic parameters. This system is used to implement the electromagnetic sensing and imaging methods in Examples 1, 2, and 3.

[0135] Finally, it should be noted that the above descriptions are merely preferred embodiments of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for sensing the electromagnetic properties of a three-dimensional target based on a reconfigurable smart surface, characterized in that, Includes the following steps: Step 1: Construct an electromagnetic sensing system consisting of a transmitting antenna, a reconfigurable smart surface (RIS), a target area, and a receiving antenna, and configure the system parameters, including the operating frequency, wavelength, RIS unit parameters, transmitting antenna position, receiving antenna position, and discrete grid parameters of the target area. Step 2: Generate an incident electric field through the transmitting antenna. With the incident magnetic field The MCBB algorithm is used to generate a scattering coefficient matrix based on mutual coherence optimization. ; Step 3: Scattering coefficient matrix A reconfigurable smart surface (RIS) is configured, and the electromagnetic waves generated by the transmitting antenna are reconstructed and scattered by the RIS to generate a RIS scattering field. The method of moments is used to solve for the distribution of the RIS scattering field on the discrete grid of the target region to obtain the discretized RIS scattering field of the target region. ; Step 4: The RIS scattered field is scattered by objects in the target area, and the receiving antenna scattered field is obtained at the receiving antenna. The received scattered field The electromagnetic properties of the target and the discretized RIS scattering field of the target region The propagation relationship from the target area to the receiving antenna is also determined by this; Step 5: Based on the scattering coefficient matrix Discretized lighting field for target area and receiving scattered fields A sensing equation is constructed between the electromagnetic properties of the target and the received scattered field, and a multi-target Bayesian compressed sensing reconstruction algorithm based on compressed sensing theory is used for reconstruction.

2. The method for sensing the electromagnetic properties of a three-dimensional target based on a reconfigurable smart surface as described in claim 1, characterized in that, Incident electric field in step 2 With the incident magnetic field The specific formula is: ； in, For free space wave impedance, The wavenumber in the background medium. This is the elevation angle of the transmitting antenna.

3. The method for sensing the electromagnetic properties of a three-dimensional target based on a reconfigurable smart surface as described in claim 1, characterized in that, In step 2, the MCBB algorithm is used to generate the scattering coefficient matrix based on mutual coherence optimization. The process includes the following steps: Step 2.1: The RIS scattered field has a linear mapping relationship during its propagation to the target region. The propagation matrix is ​​obtained by discretizing this process using the method of moments. The process is described as follows: The propagation matrix With RIS scattering coefficient matrix Transpose and assign, that is , ; Step 2.2, for the matrix Perform singular value decomposition to obtain ; Step 2.3, based on the matrix Construct the target correlation matrix The formula is ; Step 2.4, for Problem, definition ; Step 2.5: Calculate the matrix Projected to Constructing an equivalent optimization matrix in Zhang Cheng's eigenspace. ; Step 2.6: Optimize the equivalent matrix Perform eigenvalue decomposition to obtain The eigenvalues ​​are arranged in descending order of their real parts as follows: ; Step 2.7: Determine the number of valid non-negative eigenvalues. ; Step 2.8: Based on the number of measurements Number of valid eigenvalues Based on the size relationship, construct an auxiliary matrix: ； Step 2.9, according to the definition Reverse reasoning ,in For RIS power.

4. The method for sensing the electromagnetic properties of a three-dimensional target based on a reconfigurable smart surface as described in claim 1, characterized in that, Step 3 specifically includes: Approximating RIS as an infinite PEC, we remove the PEC according to the physical equivalence theorem and introduce an equivalent current density on the surface. Using the PO approximation, we get: ； in, ; The formulas for the magnetic vector potential under the Lorentz gauge and the RIS scattered electric field are as follows: ； in, Representing angular frequency, magnetic permeability, and dielectric constant of the medium, respectively, the scattered field vector of the target region... Discretize and distinguish directions as follows: ； Among them, superscript They represent Axial direction, and They represent the first The RIS scattering unit and the first There are sampling units, among which ; They represent the first The RIS scattering unit and the first The x-axis and y-axis coordinates of each sampling unit Indicates the first The RIS scattering unit and the first The distance between each sampling unit Indicates the first The second measurement of RIS The reflection coefficient of a small piece; After discretization, the three directions of the RIS scattering field are further combined into the same matrix, resulting in: ;where the matrix yes The combined discretized integral kernel of each component is represented as follows: 。 5. The method for sensing the electromagnetic properties of a three-dimensional target based on a reconfigurable smart surface as described in claim 1, characterized in that, Step 4 specifically includes: Assuming the receiving antenna scattering field To accept only The target scattered electric field component along the axis, then the electric field component Represented as: ; The three-dimensional free-space Green's function tensor is: , This represents the position vector of the receiving point. The position vector representing the center of the target discrete element. Indicates that it is located at along The unit source in the direction produced at the place Directional electric field components; Indicates calculation The Green's function operator for the axially scattered target field, where Let the volume of the target discrete element be denoted as . The total number of units in the three directions after discretizing the target region; Represents the contrast vector of the target sampling point Diagonalization, contrast is expressed as: ;in, The relative permittivity, The dielectric constant in vacuum is . For electrical conductivity, ω is the angular frequency of the electromagnetic wave; For electric field components Linearization: ； ； Among them, the contrast vector of the target sampling point , This represents the Khatri-Rao product.

6. The method for sensing the electromagnetic properties of a three-dimensional target based on a reconfigurable smart surface as described in claim 1, characterized in that, Step 5 specifically includes: Based on the contrast vector of the target sampling points The perceptual equation is expressed as: ; in, Given a set of additive Gaussian noise vectors with zero mean and variance varying with the signal-to-noise ratio, As a sparse vector, the multi-objective Bayesian compressed sensing reconstruction algorithm mt-CS based on compressed sensing theory is used to solve it. Reconstructing using norm problems: .

7. A three-dimensional target electromagnetic property sensing system for implementing the three-dimensional target electromagnetic property sensing method based on a reconfigurable smart surface according to any one of claims 1-6, comprising an antenna transmitter, a programmable metasurface, an antenna receiver, and a signal processing unit; wherein the programmable metasurface is used for phase and amplitude modulation of the transmitted wave; the antenna receiver is used for receiving the scattered field; and the signal processing unit is used for performing MCBB algorithm to design the reflection coefficient of the RIS unit, construct the sensing matrix, and solve for the electromagnetic parameters.

8. The three-dimensional target electromagnetic property sensing system as described in claim 7, characterized in that, The scattering coefficient matrix is ​​constructed using the MCBB algorithm; the scattering coefficient matrix is ​​configured on a reconfigurable smart surface to reconstruct the incident field phase and amplitude modulation; the receiving antenna is fixed at the position. Used to receive the scattered field and perform compressed sensing MT-CS algorithm inversion.

Images