Channel simulation method and device based on multi-scale roughness modeling, equipment and storage medium

Abstract

The invention discloses a channel simulation method and device based on multi-scale roughness modeling, equipment and a storage medium, and relates to the technical field of three-dimensional reconstruction, and the method comprises the steps: generating a macroscopic geometric model based on macroscopic geometric point cloud data of a target scene; carrying out adaptive subdivision on the macroscopic geometric model to generate a geometric model with mesoscopic detail features and macroscopic geometric features; based on a microscopic fractal roughness model, fusing the microscopic rough features with a geometric model with mesoscopic detail features and macroscopic geometric features to generate a multi-scale geometric model; and performing channel simulation in the target scene based on the multi-scale geometric model, and predicting the channel path loss of the target scene. By means of the mode, grid macroscopic scene representation and fractal microscopic details are fused, fractal dimensions are used for driving high-frequency rough feature construction, continuous scale transition from a macroscopic scene-level geometric shape to a microscopic fractal elevation field is achieved, the limitation of single-scale modeling is broken through, and the channel prediction precision is improved.

Description

Technical Field

[0001] This application relates to the field of three-dimensional reconstruction technology, and in particular to channel simulation methods, devices, equipment and storage media based on multi-scale roughness modeling. Background Technology

[0002] As 5G / 6G communication evolves towards millimeter-wave (mmWave) and terahertz (THz) bands (such as 28GHz, 60GHz, and even higher frequencies), the wavelength of wireless signals is shrinking to the centimeter or even millimeter scale (e.g., 30GHz corresponds to a wavelength of approximately 1cm). In this context, traditional geometric models based on oblique photography and multi-view 3D reconstruction (typically relying on smoothed and optimized meshes or point clouds) cannot accurately represent the microscopic geometric features (such as subwavelength undulations and fractal structures) of various surfaces in the real environment (such as rough walls, vegetation, and furniture). These microstructures significantly affect the scattering, diffraction, and multipath propagation behavior of high-frequency electromagnetic waves, leading to significant discrepancies between channel simulation (such as ray tracing) predictions and actual measurements. Therefore, a new method is needed that integrates high-precision 3D geometric reconstruction with electromagnetic modeling of surface roughness to improve the channel simulation accuracy of high-frequency wireless communication systems and support practical applications such as base station deployment, beam optimization, and network planning.

[0003] The above content is only used to help understand the technical solution of the present invention and does not represent an admission that the above content is prior art. Summary of the Invention

[0004] The main objective of this application is to provide a channel simulation method, apparatus, device, and storage medium based on multi-scale roughness modeling, aiming to solve the technical problem that the geometric models generated by traditional three-dimensional reconstruction methods in the prior art are difficult to meet the channel simulation requirements of high-frequency wireless communication, and the simulation prediction results have large deviations from the actual measurements.

[0005] To achieve the above objectives, this application provides a channel simulation method based on multi-scale roughness modeling, the method comprising: Generate a macro-geometric model based on macro-geometric point cloud data of the target scene; The macroscopic geometric model is adaptively subdivided to generate a geometric model with mesoscopic detail features and macroscopic geometric features; Based on the micro-fractal roughness model, the micro-roughness features are fused with the geometric model having meso-detail features and macro-geometric features to generate a multi-scale geometric model; Based on the multi-scale geometric model, channel simulation is performed in the target scenario to predict the channel path loss of the target scenario.

[0006] In one embodiment, the step of adaptively subdividing the macroscopic geometric model to generate a macroscopic geometric model with mesoscopic detail features includes: Based on the octree data structure, the macroscopic geometric model is spatially partitioned to determine the root node and the initial leaf node corresponding to the root node, and the root node covers the target scene. Obtain the number of triangular faces in the initial leaf node and the local curvature of the vertices of the triangular faces in the initial leaf node; When the number of triangular faces in the initial leaf node is greater than a preset number threshold or the local curvature of the vertices of the triangular faces in the initial leaf node is greater than a preset curvature threshold, the initial leaf node is adaptively subdivided to determine the target leaf node. Based on the root node and the target leaf node, a geometric model with mesoscopic detail features and macroscopic geometric features is generated.

[0007] In one embodiment, after the step of generating a geometric model with mesoscopic detail features and macroscopic geometric features based on the root node and the target leaf node, the method further includes: Based on the UV coordinates corresponding to the vertex coordinates of the triangular facets in the geometric model, the coordinate mapping relationship is determined and stored in the vertex attributes of the triangular facets in the geometric model. Based on the surface texture information of the target scene, a corresponding mesh texture map is generated in the geometric model, and a corresponding texture patch is loaded in the high curvature region of the target.

[0008] In one embodiment, the step of fusing micro-roughness features with the geometric model having meso-level detail features and macro-level geometric features based on the micro-fractal roughness model to generate a multi-scale geometric model includes: A micro-fractal roughness model is constructed based on a preset fractal function; Perform a two-dimensional Fourier transform on the macroscopic geometric point cloud data to determine the power spectral density; The slope corresponding to the power spectral density is fitted in a double logarithmic coordinate system, and the fractal dimension is determined based on the slope. Determine the cutoff frequency based on the target wavelength; Based on the fractal dimension, the cutoff frequency, the spatial frequency factor, and the random phase, the height field of the micro-fractal roughness model is determined. Based on the height field of the micro-fractal roughness model, the micro-roughness features are embedded into the geometric model to obtain a multi-scale geometric model.

[0009] In one embodiment, the step of performing channel simulation based on the multi-scale geometric model in the target scenario to predict the channel path loss of the target scenario includes: Based on the multi-scale geometric model, the specular reflection simulation model and diffuse scattering simulation model for the target scene are determined. Based on the aforementioned specular reflection simulation model and the aforementioned diffuse scattering simulation model, a multi-component scattering simulation model is generated. A depolarization simulation model is generated based on Stokes vectors and Mueller matrices. Based on the multi-component scattering simulation model and the depolarization simulation model, the channel path loss of the target scenario is determined.

[0010] In one embodiment, the step of determining the specular reflection simulation model and the diffuse scattering simulation model of the target scene based on the multi-scale geometric model includes: When the ray hits the macroscopic surface of the multi-scale geometric model, the height field of the micro-fractal roughness model is sampled at multiple levels in a preset local coordinate system; Based on the height field obtained from sampling, the effective normal vector and effective vertices are determined. Based on the effective normal vectors and the effective vertices, the initial normals of the multi-scale geometric model are corrected to determine the corrected normals; Based on the fractal dimension corresponding to the height field of the micro-fractal roughness model, the normal distribution function of the non-Gaussian micro-surface element is determined. Based on the corrected normal and the incident ray, a simulation model for specular reflection is determined; Based on the normal distribution function of the non-Gaussian micro-surface element, a diffuse scattering simulation model is determined.

[0011] In one embodiment, the step of determining the channel path loss of the target scenario based on the multi-component scattering simulation model and the depolarization simulation model includes: Based on the multi-component scattering simulation model, the path loss of the corrected rough surface is determined. Determine the depolarization path loss based on the depolarization simulation model; Based on the corrected rough surface path loss and the depolarization path loss, the channel path loss of the target scenario is determined.

[0012] Furthermore, to achieve the above objectives, this application also proposes a channel simulation device based on multi-scale roughness modeling, which includes: The multi-scale modeling module is used to generate macro-geometric models based on macro-geometric point cloud data of the target scene; The multi-scale modeling module is also used to adaptively subdivide the macroscopic geometric model to generate a geometric model with mesoscopic detail features and macroscopic geometric features. The multi-scale modeling module is also used to fuse micro-roughness features with the geometric model having meso-detail features and macro-geometric features based on the micro-fractal roughness model to generate a multi-scale geometric model. The simulation prediction module is used to perform channel simulation in the target scenario based on the multi-scale geometric model, and to predict the channel path loss of the target scenario.

[0013] Furthermore, to achieve the above objectives, this application also proposes a channel simulation device based on multi-scale roughness modeling. The channel simulation device based on multi-scale roughness modeling includes: a memory, a processor, and a computer program stored in the memory and executable on the processor. The computer program is configured to implement the steps of the channel simulation method based on multi-scale roughness modeling as described above.

[0014] In addition, to achieve the above objectives, the present invention also proposes a storage medium, which is a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, it implements the steps of the channel simulation method based on multi-scale roughness modeling as described above.

[0015] In addition, to achieve the above objectives, this application also provides a computer program product, which includes a computer program that, when executed by a processor, implements the steps of the channel simulation method based on multi-scale roughness modeling as described above.

[0016] This application provides a channel simulation method based on multi-scale roughness modeling. Based on macroscopic geometric point cloud data of the target scene, a macroscopic geometric model is generated. The macroscopic geometric model is adaptively subdivided to generate a geometric model with mesoscopic detail features and macroscopic geometric features. Based on a microscopic fractal roughness model, the microscopic roughness features are fused with the geometric model containing mesoscopic detail features and macroscopic geometric features to generate a multi-scale geometric model. Based on the multi-scale geometric model, channel simulation is performed in the target scene to predict the channel path loss. This application integrates the macroscopic scene representation of traditional LOD meshes with the microscopic detail generation capability of WM fractal functions. It utilizes fractal dimension to drive the construction of high-frequency roughness features, achieving a continuous scale transition from macroscopic scene-level geometric shape to WM microscopic fractal elevation field. This overcomes the limitations of single-scale modeling, improves the prediction accuracy of channel simulation in high-frequency wireless communication scenarios, and enhances the accuracy of simulation results. It solves the technical problem that the geometric models generated by traditional 3D reconstruction methods cannot meet the channel simulation requirements of high-frequency wireless communication, and that simulation prediction results deviate significantly from actual measurements. Attached Figure Description

[0017] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0018] To more clearly illustrate the technical solutions in 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, 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 an embodiment of the channel simulation method based on multi-scale roughness modeling in this application. Figure 2 This is a flowchart illustrating Embodiment 2 of the channel simulation method based on multi-scale roughness modeling in this application; Figure 3 A simplified flowchart of the channel simulation method based on multi-scale roughness modeling provided in Embodiment 2 of this application; Figure 4 This is a schematic diagram of the module structure of the channel simulation device based on multi-scale roughness modeling according to an embodiment of this application; Figure 5 This is a schematic diagram of the device structure of the hardware operating environment involved in the channel simulation method based on multi-scale roughness modeling in the embodiments of this application.

[0020] The realization of the purpose, functional features and advantages of this application will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0021] It should be understood that the specific embodiments described herein are merely illustrative of the technical solutions of this application and are not intended to limit this application.

[0022] To better understand the technical solution of this application, a detailed description will be provided below in conjunction with the accompanying drawings and specific implementation methods.

[0023] The main solution of this application embodiment is as follows: Based on the macroscopic geometric point cloud data of the target scene, a macroscopic geometric model is generated; the macroscopic geometric model is adaptively subdivided to generate a geometric model with mesoscopic detail features and macroscopic geometric features; based on the microscopic fractal roughness model, the microscopic roughness features are fused with the geometric model with mesoscopic detail features and macroscopic geometric features to generate a multi-scale geometric model; based on the multi-scale geometric model, channel simulation is performed in the target scene to predict the channel path loss of the target scene.

[0024] This application provides a solution that integrates the macroscopic scene representation of traditional LOD (Levels of Detail) meshes with the microscopic detail generation capability of WM (Weierstrass-Mandelbrot) fractal functions. By utilizing fractal dimension to drive the construction of high-frequency coarse features, it achieves a continuous scale transition from macroscopic scene-level geometry to WM microscopic fractal elevation fields. This overcomes the limitations of single-scale modeling, improves the prediction accuracy of channel simulation in high-frequency wireless communication scenarios, and enhances the accuracy of simulation results. It solves the technical problem that the geometric models generated by traditional 3D reconstruction methods cannot meet the channel simulation requirements of high-frequency wireless communication, and that there is a large deviation between simulation prediction results and actual measurements.

[0025] It should be noted that the executing entity in this embodiment can be a computing service device with data processing, network communication, and program execution functions, such as a tablet computer, personal computer, or mobile phone, or an electronic device capable of performing the above functions, a channel simulation device based on multi-scale roughness modeling, etc. This embodiment does not specifically limit it. The following uses a channel simulation device based on multi-scale roughness modeling as an example to describe this embodiment and the following embodiments.

[0026] This application provides a channel simulation method based on multi-scale roughness modeling, referring to... Figure 1 , Figure 1 This is a flowchart illustrating the first embodiment of the channel simulation method based on multi-scale roughness modeling in this application.

[0027] In this embodiment, the channel simulation method based on multi-scale roughness modeling includes steps S10~S40: Step S10: Generate a macro-geometric model based on the macro-geometric point cloud data of the target scene; It should be noted that the target scenario, i.e., the scenario requiring channel simulation, can be high-frequency wireless communication; this embodiment does not specifically limit it. Macroscopic geometric point cloud data refers to point cloud data with macroscopic geometric features, and the resolution needs to reach sub-centimeter level (e.g., 0.5cm). This embodiment uses LiDAR to acquire this data. Macroscopic geometric features can include wall outlines, furniture layouts, etc., and this embodiment does not specifically limit them.

[0028] Additionally, it should be noted that, in addition to macroscopic geometric point cloud data, this embodiment also needs to collect surface texture information. In specific implementation, a high-resolution camera (≥20 million pixels) can be used for multi-view photography, covering the visible light band (400-700nm), to obtain surface texture information.

[0029] In one feasible implementation, step S10 may include: determining sparse point cloud data of the target scene based on macroscopic geometric point cloud data of the target scene; densifying the sparse point cloud data to generate a triangular mesh model, using the triangular mesh model as the macroscopic geometric model; and calculating the surface normal vector, curvature distribution, and spatial topological relationship of the macroscopic geometric model.

[0030] Understandably, this embodiment uses the SfM (Structure from Motion) algorithm to process the image sequence (macro-geometric point cloud data), recovering the camera pose and sparse point cloud of the scene to obtain sparse point cloud data of the target scene. Then, the MVS (Multi-View Stereo) algorithm is used to densify the sparse point cloud data, generating a triangular mesh model. This triangular mesh model serves as the macro-geometric model, with the mesh side length controlled between 1-5 cm to meet the requirements of macro-electromagnetic simulation. Finally, based on the mesh, surface normal vectors are calculated. The curvature distribution and spatial topological relationship provide a benchmark for subsequent roughness superposition.

[0031] It should be understood that, for the macroscopic geometry layer (LOD0), this embodiment constructs a macroscopic geometry model and calculates data such as the macroscopic surface curvature, normal vector, and spatial relationships of the scene.

[0032] Step S20: Adaptively subdivide the macroscopic geometric model to generate a geometric model with mesoscopic detail features and macroscopic geometric features; It should be noted that for the meso-geometric layer (LOD1), adaptive subdivision is performed based on an octree, retaining features of 1mm to 1cm.

[0033] In one feasible implementation, step S20 may include steps S201 to S204: Step S201: Based on the octree data structure, the macroscopic geometric model is spatially partitioned to determine the root node and the initial leaf node corresponding to the root node, wherein the root node covers the target scene. It should be noted that an octree data structure is used to spatially partition the macroscopic triangular mesh (LOD0). The root node covers the entire 3D scene bounding box. Child nodes (leaf nodes) are generated through recursive partitioning (up to 8 levels), and the resolution of each leaf node increases in powers of 2. The initial leaf node is the leaf node obtained from the preliminary partitioning, for example, the first layer leaf node. Further determination is needed to determine whether further partitioning is required.

[0034] It is understood that in this embodiment, the resolution of the root node is half the length of the longest side in the target scene, and the minimum resolution of the leaf node is 1mm (corresponding to the terahertz band) to 1cm (corresponding to the millimeter wave band). In specific implementation, it can be adjusted according to the wavelength of the target signal (e.g., 1cm for the 30GHz band).

[0035] Step S202: Obtain the number of triangular faces in the initial leaf node and the local curvature of the vertices of the triangular faces in the initial leaf node; Understandably, if the number of triangular faces contained in a node exceeds the threshold (default is 50 per node) or the local curvature is greater than the set threshold (default is 0.05 / m), subdivision is triggered until the termination condition is met (i.e., the set resolution or the set number of faces is reached).

[0036] Step S203: When the number of triangular faces in the initial leaf node is greater than a preset number threshold or the local curvature of the vertices of the triangular faces in the initial leaf node is greater than a preset curvature threshold, the initial leaf node is adaptively subdivided to determine the target leaf node. It should be noted that the preset quantity threshold is the threshold for the number of triangular facets, which is usually set to 50, and the preset curvature threshold is the threshold for the local curvature, which is usually set to 0.05 / m.

[0037] Understandably, if the number of triangles contained in a node exceeds a preset threshold or the local curvature of a vertex is greater than a preset curvature threshold, subdivision is triggered until the number of triangles is less than or equal to the preset threshold or the local curvature of a vertex is less than or equal to the preset curvature threshold.

[0038] It should be understood that for each vertex of a triangular facet, the local curvature is estimated using a least-squares plane fitting method, with a curvature threshold T (e.g., 0.05 / m). When the local curvature (|H|) of a vertex is greater than T, it is marked as a "feature vertex," triggering adaptive subdivision of the node. In the triangular facet containing the feature vertex, a new vertex is inserted along the midpoint of the longest edge, splitting the original facet into two sub-facets, until the edge length of the neighborhood mesh of all feature vertices is reached. After subdivision, perform a Laplace operation on the non-boundary vertices:

[0039] in, For non-boundary vertices, For vertices that have undergone the Laplace operation, a total of 3 iterations are required, with a smoothing factor. This is used to reduce the high-frequency noise introduced by subdivision.

[0040] Step S204: Based on the root node and the target leaf node, generate a geometric model with mesoscopic detail features and macroscopic geometric features.

[0041] It is understandable that the target leaf node is the final determined leaf node. Based on the root node and target leaf node of the division, the macroscopic geometric model after mesoscopic subdivision can be obtained, that is, a geometric model with mesoscopic detailed features and macroscopic geometric features.

[0042] In one feasible implementation, step S204 may include: determining a coordinate mapping relationship based on the UV coordinates corresponding to the vertex coordinates of the triangular facets in the geometric model, storing the coordinate mapping relationship in the vertex attributes of the triangular facets in the geometric model; generating a corresponding mesh texture map in the geometric model based on the surface texture information of the target scene, and loading the corresponding texture patch in the target high curvature region.

[0043] Understandably, global UV unwrapping is performed on the subdivided mesh, using a conformal mapping method based on mesh flow to map the 3D mesh surface to a 2D UV space. The objective function is:

[0044] in, It is a grid edge set. It is a 3D Euclidean distance. UV distance, , where is the scaling factor for the side length. The solution is obtained using the conjugate gradient method, with energy change as the convergence condition. A one-to-one correspondence is established between vertex coordinates (x, y, z) and UV coordinates (u, v), i.e., the coordinate mapping relationship, stored in the mesh vertex attributes, supporting subsequent local coordinate positioning in the fractal height field (LOD2).

[0045] It should be noted that the target high curvature region refers to the region with a relatively high curvature, such as the neighborhood marked as "feature vertex". The mesh texture map is the high dynamic range (HDR) texture map of the 3D mesh.

[0046] It should be understood that by fusing visible light and near-infrared images and eliminating camera viewpoint bias through image registration algorithms (such as SIFT feature matching), HDR texture maps are generated. The texture resolution is set to at least 16×16 pixels for each triangular facet of the mesoscopic mesh, ensuring that 1mm-level features are clearly visible in the texture. For areas with high curvature, high-resolution texture patches are automatically loaded, and pyramid texture mapping technology is used to achieve a smooth transition between different levels of detail, avoiding texture seams.

[0047] Step S30: Based on the micro fractal roughness model, the micro roughness features are fused with the geometric model having meso-detail features and macro-geometric features to generate a multi-scale geometric model. It should be noted that for the micro-geometry layer (LOD2), a WM fractal height field is constructed and superimposed using normal perturbation or displacement mapping.

[0048] In one feasible implementation, step S30 may include steps S301 to S305: Step S301: Construct a micro-fractal roughness model based on a preset fractal function; It should be noted that the preset fractal function is the WM fractal function, and a subwavelength-level surface undulation model, i.e., a microscopic fractal roughness model, is constructed based on the WM fractal function. The WM fractal function can generate multi-scale self-similar rough surfaces by adjusting the fractal dimension and spatial frequency scaling factor.

[0049] Step S302: Perform a two-dimensional Fourier transform on the macroscopic geometric point cloud data to determine the power spectral density, fit the slope corresponding to the power spectral density in a double logarithmic coordinate system, and determine the fractal dimension based on the slope. Perform a two-dimensional Fourier transform on the macroscopic geometric point cloud data to calculate the power spectral density. The calculation relationship is shown below:

[0050] In the formula, This represents the power spectral density. The slope is fitted in a log-log coordinate system. Through formula The fractal dimension is obtained. , .

[0051] Step S303: Determine the cutoff frequency based on the target wavelength; It should be noted that the cutoff frequency satisfies ,in, For the target wavelength (30GHz) (0.01m) The cutoff frequency, Spatial frequency factor ( ),Pick ,at this time .

[0052] Step S304: Determine the height field of the micro-fractal roughness model based on the fractal dimension, the cutoff frequency, the spatial frequency factor, and the random phase. Understandably, the expression for the height field of the micro-fractal roughness model is:

[0053] in, For fractal dimensions, Spatial frequency factor, The cutoff frequency, It is a random phase used to introduce randomness into the surface morphology.

[0054] Step S305: Based on the height field of the micro-fractal roughness model, embed the micro-roughness features into the geometric model to obtain a multi-scale geometric model; It should be noted that this embodiment uses two methods for geometric blending: normal perturbation mapping and displacement mapping, both of which are saved through textures. Microscopic roughness features can be the roughness of concrete, the fractal structure of vegetation, etc., and this embodiment does not specifically limit them.

[0055] Understandably, if a normal perturbation map is used, then in the local coordinate system, the WM height field will be affected. By step size Sampling, gradient calculation via central difference:

[0056] Next, the gradient vector Normalized to [-1,1] and encoded as RGB channels of the normal map:

[0057] Then, during ray tracing, the perturbation vector is obtained through texture sampling, and the effective normal vector is calculated. :

[0058] It should be understood that if displacement mapping is used, the height field will be... Normalized to [0,1], stored as a 16-bit single-channel texture, vertex displacement is implemented through a geometry shader, and the calculation formula is as follows:

[0059] in, This represents the position of the vertex after displacement. The vertex position before displacement. For the surface normal vector, With a scaling factor of 1 μm / unit, the displacement accuracy reaches ±0.1 μm, and it supports real-time adjustment of the roughness amplitude.

[0060] Step S40: Based on the multi-scale geometric model, perform channel simulation in the target scenario to predict the channel path loss of the target scenario.

[0061] It should be noted that the channel path loss is the result obtained from the simulation. This embodiment utilizes a multi-scale geometric model combined with an improved ray tracing algorithm based on fractal surfaces to achieve channel simulation of the target scene, thus obtaining accurate simulation results.

[0062] This embodiment provides a channel simulation method based on multi-scale roughness modeling. Based on macroscopic geometric point cloud data of the target scene, a macroscopic geometric model is generated. The macroscopic geometric model is adaptively subdivided to generate a geometric model with mesoscopic detail features and macroscopic geometric features. Based on a microscopic fractal roughness model, the microscopic roughness features are fused with the geometric model containing mesoscopic detail features and macroscopic geometric features to generate a multi-scale geometric model. Based on the multi-scale geometric model, channel simulation is performed in the target scene to predict the channel path loss. This embodiment integrates the macroscopic scene representation of traditional LOD meshes with the microscopic detail generation capability of WM fractal functions. It utilizes fractal dimension to drive the construction of high-frequency roughness features, achieving a continuous scale transition from macroscopic scene-level geometric shape to WM microscopic fractal elevation field. This overcomes the limitations of single-scale modeling, improves the prediction accuracy of channel simulation in high-frequency wireless communication scenarios, and enhances the accuracy of simulation results.

[0063] Based on the first embodiment of this application, in the second embodiment of this application, the content that is the same as or similar to that in Embodiment 1 above can be referred to the above description, and will not be repeated hereafter. Based on this, please refer to... Figure 2 Step S40 may include steps S401 to S404: Step S401: Based on the multi-scale geometric model, determine the specular reflection simulation model and the diffuse scattering simulation model for the target scene; In one feasible implementation, step S401 may include steps A11 to A15: Step A11: When the ray hits the macroscopic surface of the multi-scale geometric model, the height field of the micro-fractal roughness model is sampled at multiple levels in a preset local coordinate system. It should be noted that when the ray hits the macroscopic surface, the LOD2 height field is sampled at multiple levels in the local coordinate system. Specifically, for the low-frequency component ( ), using step size Perform uniform sampling to ensure the Nyquist frequency condition; for high-frequency components ( Using jittered sampling to reduce aliasing, the sampling point offset... .

[0064] Step A12: Based on the height field obtained from sampling, determine the effective normal vector and effective vertices; Understandably, the effective normal vector and effective vertices are calculated using the height field obtained from sampling through the central difference method.

[0065] Step A13: Based on the effective normal vector and the effective vertex, the initial normal of the multi-scale geometric model is modified to determine the modified normal; It is understandable that the normal is modified by using effective normal vectors and effective vertices to represent microscopic undulations, and the modified normal is called the modified normal.

[0066] Step A14: Determine the normal distribution function of the non-Gaussian micro-surface element based on the fractal dimension corresponding to the height field of the micro-fractal roughness model; It should be noted that, based on the fractal dimension, the non-Gaussian micro-element normal distribution function (NDF) is defined as follows:

[0067] in, For gamma function, The angle between the micro-element normal and the macro-element normal. The dispersion of the height of each sampling point on the surface relative to the average height is an important indicator for measuring surface roughness. This distribution is... It degenerates into a Gaussian distribution at time, It tends to approximate an isotropic coarse distribution.

[0068] Step A15: Based on the corrected normal and the incident ray, determine the specular reflection simulation model; based on the non-Gaussian micro-surface element normal distribution function, determine the diffuse scattering simulation model.

[0069] It is understood that the ray tracing algorithm in this embodiment mainly considers two parts: the specular reflection component and the diffuse scattering component. For the specular reflection component, given the phase-corrected normal and the incident ray, the formula for calculating the reflection direction is:

[0070] In the formula, To correct the normal, For the incident light ray, The reflection direction is defined. Based on the calculation formula for the reflection direction, a specular reflection simulation model is constructed.

[0071] It should be understood that, for the diffuse scattering component, importance sampling is first used from... Generated in Normal vectors of micro-element The sampling formula is:

[0072] In the formula, , The polar angle represents the angle between the polar angle and the macroscopic normal. Let be the direction angle. A diffuse scattering normal can be determined from the polar angle and the direction angle:

[0073] For each Calculate the scattering direction:

[0074] In the formula, The scattering direction, For the incident light ray, Let be the normal vector of the micro-surface element. Based on the calculation formula for the scattering direction, a diffuse scattering simulation model is constructed.

[0075] Step S402: Based on the specular reflection simulation model and the diffuse scattering simulation model, generate a multi-component scattering simulation model; It should be noted that a multi-component scattering simulation model is generated by combining the specular reflection simulation model and the diffuse scattering simulation model.

[0076] It is understandable that by introducing a surface micro-perturbation field... With macroscopic incident normal Effective normal formed by coupling By combining the distribution function of the normal direction of non-Gaussian micro-surface elements, a multi-component scattering simulation model based on importance sampling is constructed. This allows for the control of the influence of fractal dimension on the diffuse scattering direction, enabling physical-precision simulation of surface-induced non-mirror scattering, and significantly improving the modeling capability and prediction accuracy of high-frequency ray tracing systems in complex and rough target scenarios.

[0077] Step S403: Generate a depolarization simulation model based on Stokes vectors and Mueller matrices; It should be noted that in high-frequency wireless communication scenarios, the depolarization effect refers to the phenomenon that the polarization state changes after electromagnetic waves are scattered by a rough surface (such as the generation of cross-polarization components in linearly polarized waves). The asymmetry and multi-scale characteristics of fractal surfaces significantly enhance the depolarization effect; therefore, a polarization-related scattering model needs to be introduced into the ray tracing algorithm.

[0078] It is understandable that the polarization state of an electromagnetic wave can be described by the Stokes vector, as shown below:

[0079] in, Total light intensity The difference in horizontal-vertical polarization intensity The polarization intensity difference is ±45°. The difference in right-handed versus left-handed circular polarization intensity This is a Stokes vector.

[0080] It should be understood that this embodiment uses the Mueller matrix to describe the polarization transformation, as shown below:

[0081] In the formula, The Mueller matrix can be obtained by integrating the Jones matrix. For the initial Stokes vector, This is the transformed Stokes vector.

[0082] Step S404: Based on the multi-component scattering simulation model and the depolarization simulation model, determine the channel path loss of the target scenario.

[0083] In one feasible implementation, step S404 may include steps B11-B13: Step B11: Based on the multi-component scattering simulation model, determine the path loss of the corrected rough surface. It is understandable that the corrected rough surface path loss is the rough surface path loss after correction, and the calculation relationship is as follows:

[0084] in, The root mean square of the surface height is used to measure surface roughness (for fractal rough surfaces, this can be achieved through the fractal dimension). and related length Derivation, ), It is the electromagnetic wave incident angle (the angle between the signal and the surface normal when the signal is incident on the rough surface). The wavelength of electromagnetic waves. To correct path loss on rough surfaces, For smooth surface path loss.

[0085] Step B12: Determine the depolarization path loss based on the depolarization simulation model; It is understood that this embodiment also considers the path loss caused by the depolarization effect of the rough surface, i.e., the depolarization path loss, and the calculation relationship is shown below:

[0086] In the formula, For depolarization path loss, , , , For Muller matrix, , , , represents the element value in the Stokes vector corresponding to the electromagnetic wave.

[0087] Step B13: Based on the corrected rough surface path loss and the depolarization path loss, determine the channel path loss of the target scenario.

[0088] It is understandable that the formula for calculating the channel path loss in the target scenario is:

[0089] In the formula, For channel path loss, For depolarization path loss, To correct path loss on rough surfaces.

[0090] This embodiment provides a channel simulation method based on multi-scale roughness modeling. Based on a multi-scale geometric model, it determines the specular reflection simulation model and diffuse scattering simulation model for the target scene. Based on the specular reflection and diffuse scattering simulation models, it generates a multi-component scattering simulation model. Based on Stokes vectors and Mueller matrices, it generates a depolarization simulation model. Based on the multi-component scattering and depolarization simulation models, it determines the channel path loss for the target scene. This embodiment calculates path loss based on a geometry-electromagnetic coupling mechanism, which breaks the simplistic assumption of ideal smooth surfaces in traditional models, improves the accuracy of path loss estimation, and supports fine-grained propagation simulation and channel modeling of high-frequency signals in complex scenes.

[0091] For example, to help understand the implementation flow of the channel simulation method based on multi-scale roughness modeling obtained by combining this embodiment with the above-described embodiment two, please refer to... Figure 3 , Figure 3 A simplified flowchart of a channel simulation method based on multi-scale roughness modeling is provided, specifically: Geometric and microscopic representation of a three-dimensional rough surface: WM fractal functions can adjust the fractal dimension. and spatial frequency scaling factor A multi-scale self-similar rough surface is generated. Its height field... It can be represented as:

[0092] in, The fractal dimension (estimated by laser scanning or image analysis). Typically, a cutoff frequency of 1.5 is used. The minimum wavelength is determined by the electromagnetic wavelength of the target frequency band (e.g., 30GHz corresponds to the minimum wavelength). ≈1cm). The generated microscopic details (μm~mm level) can be fused with macroscopic 3D reconstructed geometry (Mesh or point cloud) to form a multi-scale geometric model.

[0093] Multi-scale geometric model: Macro geometry layer (LOD0), derived from photogrammetry / SfM triangular mesh, calculates macroscopic surface curvature, normal vectors, spatial relationships, etc. of the scene; Meso geometry layer (LOD1), based on octree adaptive subdivision, retains 1mm~1cm features; Micro geometry layer (LOD2), WM fractal height field, superimposed by normal perturbation or displacement mapping.

[0094] Ray tracing based on WM fractal surfaces: Traditional ray tracing assumes smooth surfaces, while WM fractal surfaces induce non-reflective scattering. The improved ray tracing algorithm includes the following key steps: 1. Ray-surface interaction determination. When a ray hits a macroscopic surface, the WM height field is sampled in the local coordinate system, and the effective normal vector is calculated. ,in For macro normals, The gradient of the WM height field.

[0095] 2. Scattering direction sampling. Based on the normal distribution of micro-surface elements. Secondary ray directions are randomly generated to simulate diffuse scattering; for the mirror component, phase-corrected reflection directions are used.

[0096] 3. Path Loss Calculation. A roughness correction factor is introduced to calculate the corrected path loss, and the path loss caused by the depolarization effect of the rough surface is also calculated. The total path loss is calculated using the corrected rough surface path loss and the depolarization path loss.

[0097] It should be noted that the above examples are only for understanding this application and do not constitute a limitation on the channel simulation method based on multi-scale roughness modeling in this application. Any simple transformations based on this technical concept are within the protection scope of this application.

[0098] This application also provides a channel simulation device based on multi-scale roughness modeling, please refer to... Figure 4 The channel simulation device based on multi-scale roughness modeling includes: The multi-scale modeling module 10 is used to generate a macro-geometric model based on the macro-geometric point cloud data of the target scene; The multi-scale modeling module 10 is also used to adaptively subdivide the macroscopic geometric model to generate a geometric model with mesoscopic detail features and macroscopic geometric features. The multi-scale modeling module 10 is also used to fuse micro-roughness features with the geometric model having meso-detail features and macro-geometric features based on the micro-fractal roughness model to generate a multi-scale geometric model. The simulation prediction module 20 is used to perform channel simulation in the target scenario based on the multi-scale geometric model and predict the channel path loss of the target scenario.

[0099] In one feasible implementation, the multi-scale modeling module 10 is further configured to spatially partition the macroscopic geometric model based on an octree data structure, determine the root node and the initial leaf node corresponding to the root node, wherein the root node covers the target scene. Obtain the number of triangular faces in the initial leaf node and the local curvature of the vertices of the triangular faces in the initial leaf node; When the number of triangular faces in the initial leaf node is greater than a preset number threshold or the local curvature of the vertices of the triangular faces in the initial leaf node is greater than a preset curvature threshold, the initial leaf node is adaptively subdivided to determine the target leaf node. Based on the root node and the target leaf node, a geometric model with mesoscopic detail features and macroscopic geometric features is generated.

[0100] In one feasible implementation, the multi-scale modeling module 10 is further configured to determine a coordinate mapping relationship based on the UV coordinates corresponding to the vertex coordinates of the triangular facets in the geometric model, and store the coordinate mapping relationship in the vertex attributes of the triangular facets in the geometric model. Based on the surface texture information of the target scene, a corresponding mesh texture map is generated in the geometric model, and a corresponding texture patch is loaded in the high curvature region of the target.

[0101] In one feasible implementation, the multi-scale modeling module 10 is also used to construct a micro-fractal roughness model based on a preset fractal function; Perform a two-dimensional Fourier transform on the macroscopic geometric point cloud data to determine the power spectral density; The slope corresponding to the power spectral density is fitted in a double logarithmic coordinate system, and the fractal dimension is determined based on the slope. Determine the cutoff frequency based on the target wavelength; Based on the fractal dimension, the cutoff frequency, the spatial frequency factor, and the random phase, the height field of the micro-fractal roughness model is determined. Based on the height field of the micro-fractal roughness model, the micro-roughness features are embedded into the geometric model to obtain a multi-scale geometric model.

[0102] In one feasible implementation, the simulation prediction module 20 is further configured to determine the specular reflection simulation model and the diffuse scattering simulation model of the target scene based on the multi-scale geometric model. Based on the aforementioned specular reflection simulation model and the aforementioned diffuse scattering simulation model, a multi-component scattering simulation model is generated. A depolarization simulation model is generated based on Stokes vectors and Mueller matrices. Based on the multi-component scattering simulation model and the depolarization simulation model, the channel path loss of the target scenario is determined.

[0103] In one feasible implementation, the simulation prediction module 20 is further configured to perform multi-level sampling of the height field of the micro-fractal roughness model in a preset local coordinate system when the ray hits the macroscopic surface of the multi-scale geometric model. Based on the height field obtained from sampling, the effective normal vector and effective vertices are determined. Based on the effective normal vectors and the effective vertices, the initial normals of the multi-scale geometric model are corrected to determine the corrected normals; Based on the fractal dimension corresponding to the height field of the micro-fractal roughness model, the normal distribution function of the non-Gaussian micro-surface element is determined. Based on the corrected normal and the incident ray, a simulation model for specular reflection is determined; Based on the normal distribution function of the non-Gaussian micro-surface element, a diffuse scattering simulation model is determined.

[0104] In one feasible implementation, the simulation prediction module 20 is further configured to determine the path loss of the corrected rough surface based on the multi-component scattering simulation model. Determine the depolarization path loss based on the depolarization simulation model; Based on the corrected rough surface path loss and the depolarization path loss, the channel path loss of the target scenario is determined.

[0105] The channel simulation device based on multi-scale roughness modeling provided in this application, employing the channel simulation method based on multi-scale roughness modeling in the above embodiments, can solve the technical problem that the geometric models generated by traditional three-dimensional reconstruction methods are difficult to meet the channel simulation requirements of high-frequency wireless communication, and that the simulation prediction results deviate significantly from the actual measurements. Compared with the prior art, the beneficial effects of the channel simulation device based on multi-scale roughness modeling provided in this application are the same as those of the channel simulation method based on multi-scale roughness modeling provided in the above embodiments, and other technical features in the channel simulation device based on multi-scale roughness modeling are the same as those disclosed in the methods of the above embodiments, and will not be repeated here.

[0106] This application provides a channel simulation device based on multi-scale roughness modeling. The channel simulation device based on multi-scale roughness modeling includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to execute the channel simulation method based on multi-scale roughness modeling in the above embodiment 1.

[0107] The following is for reference. Figure 5 This document illustrates a structural schematic diagram of a channel simulation device based on multi-scale roughness modeling suitable for implementing embodiments of this application. The channel simulation device based on multi-scale roughness modeling in the embodiments of this application may include, but is not limited to, mobile terminals such as mobile phones, laptops, digital broadcast receivers, PDAs (Personal Digital Assistants), PADs (Portable Application Description), PMPs (Portable Media Players), and in-vehicle terminals (e.g., in-vehicle navigation terminals), as well as fixed terminals such as digital TVs and desktop computers. Figure 5 The channel simulation device based on multi-scale roughness modeling shown is merely an example and should not impose any limitations on the functionality and scope of use of the embodiments of this application.

[0108] like Figure 5As shown, the channel simulation device based on multi-scale roughness modeling may include a processing unit 1001 (e.g., a central processing unit, a graphics processing unit, etc.), which can perform various appropriate actions and processes according to a program stored in ROM (Read Only Memory) 1002 or a program loaded from storage device 1003 into RAM (Random Access Memory) 1004. RAM 1004 also stores various programs and data required for the operation of the channel simulation device based on multi-scale roughness modeling. The processing unit 1001, ROM 1002, and RAM 1004 are interconnected via bus 1005. Input / output (I / O) interface 1006 is also connected to the bus. Typically, the following systems can be connected to I / O interface 1006: input devices 1007 including, for example, touchscreens, touchpads, keyboards, mice, image sensors, microphones, accelerometers, gyroscopes, etc.; output devices 1008 including, for example, liquid crystal displays (LCDs), speakers, vibrators, etc.; storage devices 1003 including, for example, magnetic tapes, hard disks, etc.; and communication devices 1009. Communication device 1009 allows the channel simulation device based on multi-scale roughness modeling to communicate wirelessly or wiredly with other devices to exchange data. Although a channel simulation device based on multi-scale roughness modeling with various systems is shown in the figure, it should be understood that it is not required to implement or possess all the systems shown. More or fewer systems can be implemented alternatively.

[0109] Specifically, according to the embodiments disclosed in this application, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments disclosed in this application include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via a communication device, or installed from storage device 1003, or installed from ROM 1002. When the computer program is executed by processing device 1001, it performs the functions defined in the methods of the embodiments disclosed in this application.

[0110] The channel simulation device based on multi-scale roughness modeling provided in this application, employing the channel simulation method based on multi-scale roughness modeling in the above embodiments, can solve the technical problem that the geometric models generated by traditional three-dimensional reconstruction methods are difficult to meet the channel simulation requirements of high-frequency wireless communication, and that the simulation prediction results deviate significantly from the actual measurements. Compared with the prior art, the beneficial effects of the channel simulation device based on multi-scale roughness modeling provided in this application are the same as those of the channel simulation method based on multi-scale roughness modeling provided in the above embodiments, and other technical features in this channel simulation device based on multi-scale roughness modeling are the same as those disclosed in the previous embodiment method, and will not be repeated here.

[0111] It should be understood that the various parts disclosed in this application can be implemented using hardware, software, firmware, or a combination thereof. In the description of the above embodiments, specific features, structures, materials, or characteristics can be combined in any suitable manner in one or more embodiments or examples.

[0112] The above are merely specific embodiments of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0113] This application provides a computer-readable storage medium having computer-readable program instructions (i.e., a computer program) stored thereon, which are used to execute the channel simulation method based on multi-scale roughness modeling in the above embodiments.

[0114] The computer-readable storage medium provided in this application may be, for example, a USB flash drive, but is not limited to, electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices, or any combination thereof. More specific examples of computer-readable storage media may include, but are not limited to: electrical connections having one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof. In this embodiment, the computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, system, or device. The program code contained on the computer-readable storage medium may be transmitted using any suitable medium, including but not limited to: wires, optical cables, RF (Radio Frequency), etc., or any suitable combination thereof.

[0115] The aforementioned computer-readable storage medium may be included in a channel simulation device based on multi-scale roughness modeling; or it may exist independently and not assembled into a channel simulation device based on multi-scale roughness modeling.

[0116] The aforementioned computer-readable storage medium carries one or more programs. When these programs are executed by a channel simulation device based on multi-scale roughness modeling, the channel simulation device based on multi-scale roughness modeling: generates a macroscopic geometric model based on macroscopic geometric point cloud data of the target scene; adaptively subdivides the macroscopic geometric model to generate a geometric model with mesoscopic detail features and macroscopic geometric features; based on a microscopic fractal roughness model, fuses the microscopic roughness features with the geometric model with mesoscopic detail features and macroscopic geometric features to generate a multi-scale geometric model; and based on the multi-scale geometric model, performs channel simulation in the target scene to predict the channel path loss of the target scene.

[0117] Computer program code for performing the operations of this application can be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, and C++, and conventional procedural programming languages ​​such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network—including a Local Area Network (LAN) or a Wide Area Network (WAN)—or can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0118] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of this application. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0119] The modules described in the embodiments of this application can be implemented in software or hardware. The names of the modules do not necessarily limit the functionality of the unit itself.

[0120] The readable storage medium provided in this application is a computer-readable storage medium that stores computer-readable program instructions (i.e., a computer program) for executing the aforementioned channel simulation method based on multi-scale roughness modeling. This solves the technical problem that the geometric models generated by traditional 3D reconstruction methods are insufficient to meet the channel simulation requirements of high-frequency wireless communication, and that simulation prediction results deviate significantly from actual measurements. Compared with the prior art, the beneficial effects of the computer-readable storage medium provided in this application are the same as those of the channel simulation method based on multi-scale roughness modeling provided in the above embodiments, and will not be repeated here.

[0121] This application also provides a computer program product, including a computer program that, when executed by a processor, implements the steps of the channel simulation method based on multi-scale roughness modeling as described above.

[0122] The computer program product provided in this application can solve the technical problem that the geometric models generated by traditional 3D reconstruction methods are difficult to meet the channel simulation requirements of high-frequency wireless communication, and the simulation prediction results deviate significantly from the actual measurements. Compared with the prior art, the beneficial effects of the computer program product provided in this application are the same as those of the channel simulation method based on multi-scale roughness modeling provided in the above embodiments, and will not be repeated here.

[0123] The above are only some embodiments of this application and do not limit the patent scope of this application. All equivalent structural transformations made under the technical concept of this application and using the contents of the specification and drawings of this application, or direct / indirect applications in other related technical fields, are included in the patent protection scope of this application.

Claims

1. A channel simulation method based on multi-scale roughness modeling, characterized in that, The method includes: Generate a macro-geometric model based on macro-geometric point cloud data of the target scene; The macroscopic geometric model is adaptively subdivided to generate a geometric model with mesoscopic detail features and macroscopic geometric features; Based on the micro-fractal roughness model, the micro-roughness features are fused with the geometric model having meso-detail features and macro-geometric features to generate a multi-scale geometric model; Based on the multi-scale geometric model, channel simulation is performed in the target scenario to predict the channel path loss of the target scenario.

2. The method as described in claim 1, characterized in that, The step of adaptively subdividing the macroscopic geometric model to generate a macroscopic geometric model with mesoscopic detail features includes: Based on the octree data structure, the macroscopic geometric model is spatially partitioned to determine the root node and the initial leaf node corresponding to the root node, and the root node covers the target scene. Obtain the number of triangular faces in the initial leaf node and the local curvature of the vertices of the triangular faces in the initial leaf node; When the number of triangular faces in the initial leaf node is greater than a preset number threshold or the local curvature of the vertices of the triangular faces in the initial leaf node is greater than a preset curvature threshold, the initial leaf node is adaptively subdivided to determine the target leaf node. Based on the root node and the target leaf node, a geometric model with mesoscopic detail features and macroscopic geometric features is generated.

3. The method as described in claim 2, characterized in that, The step of generating a geometric model with mesoscopic detail features and macroscopic geometric features based on the root node and the target leaf node further includes: Based on the UV coordinates corresponding to the vertex coordinates of the triangular facets in the geometric model, the coordinate mapping relationship is determined and stored in the vertex attributes of the triangular facets in the geometric model. Based on the surface texture information of the target scene, a corresponding mesh texture map is generated in the geometric model, and a corresponding texture patch is loaded in the high curvature region of the target.

4. The method as described in claim 1, characterized in that, The step of fusing micro-fractal roughness models with geometric models possessing meso-level detail features and macro-level geometric features to generate multi-scale geometric models includes: A micro-fractal roughness model is constructed based on a preset fractal function; Perform a two-dimensional Fourier transform on the macroscopic geometric point cloud data to determine the power spectral density; The slope corresponding to the power spectral density is fitted in a double logarithmic coordinate system, and the fractal dimension is determined based on the slope. Determine the cutoff frequency based on the target wavelength; Based on the fractal dimension, the cutoff frequency, the spatial frequency factor, and the random phase, the height field of the micro-fractal roughness model is determined. Based on the height field of the micro-fractal roughness model, the micro-roughness features are embedded into the geometric model to obtain a multi-scale geometric model.

5. The method as described in claim 1, characterized in that, The step of performing channel simulation in the target scenario based on the multi-scale geometric model and predicting the channel path loss of the target scenario includes: Based on the multi-scale geometric model, the specular reflection simulation model and diffuse scattering simulation model for the target scene are determined. Based on the aforementioned specular reflection simulation model and the aforementioned diffuse scattering simulation model, a multi-component scattering simulation model is generated. A depolarization simulation model is generated based on Stokes vectors and Mueller matrices. Based on the multi-component scattering simulation model and the depolarization simulation model, the channel path loss of the target scenario is determined.

6. The method as described in claim 5, characterized in that, The steps for determining the specular reflection simulation model and diffuse scattering simulation model of the target scene based on the multi-scale geometric model include: When the ray hits the macroscopic surface of the multi-scale geometric model, the height field of the micro-fractal roughness model is sampled at multiple levels in a preset local coordinate system; Based on the height field obtained from sampling, the effective normal vector and effective vertices are determined. Based on the effective normal vectors and the effective vertices, the initial normals of the multi-scale geometric model are corrected to determine the corrected normals; Based on the fractal dimension corresponding to the height field of the micro-fractal roughness model, the normal distribution function of the non-Gaussian micro-surface element is determined. Based on the corrected normal and the incident ray, a simulation model for specular reflection is determined; Based on the normal distribution function of the non-Gaussian micro-surface element, a diffuse scattering simulation model is determined.

7. The method as described in claim 5, characterized in that, The step of determining the channel path loss of the target scenario based on the multi-component scattering simulation model and the depolarization simulation model includes: Based on the multi-component scattering simulation model, the path loss of the corrected rough surface is determined. Determine the depolarization path loss based on the depolarization simulation model; Based on the corrected rough surface path loss and the depolarization path loss, the channel path loss of the target scenario is determined.

8. A channel simulation device based on multi-scale roughness modeling, characterized in that, The device includes: The multi-scale modeling module is used to generate macro-geometric models based on macro-geometric point cloud data of the target scene; The multi-scale modeling module is also used to adaptively subdivide the macroscopic geometric model to generate a geometric model with mesoscopic detail features and macroscopic geometric features. The multi-scale modeling module is also used to fuse micro-roughness features with the geometric model having meso-detail features and macro-geometric features based on the micro-fractal roughness model to generate a multi-scale geometric model. The simulation prediction module is used to perform channel simulation in the target scenario based on the multi-scale geometric model, and to predict the channel path loss of the target scenario.

9. A channel simulation device based on multi-scale roughness modeling, characterized in that, The device includes: a memory, a processor, and a computer program stored in the memory and executable on the processor, the computer program being configured to implement the steps of the channel simulation method based on multi-scale roughness modeling as described in any one of claims 1 to 7.

10. A storage medium, characterized in that, The storage medium is a computer-readable storage medium, and a computer program is stored on the storage medium. When the computer program is executed by a processor, it implements the steps of the channel simulation method based on multi-scale roughness modeling as described in any one of claims 1 to 7.

Images