City wireless signal field strength prediction method and system based on three-dimensional scene modeling

Abstract

The application relates to a city wireless signal field strength prediction method and system based on three-dimensional scene modeling, wherein the method comprises the following steps: constructing a three-dimensional digital model according to city geographic information data, configuring parameters of the three-dimensional digital model, and generating a wireless signal simulation model; performing field strength calculation on preset sampling points based on the wireless signal simulation model, and generating a true value data set for training; constructing a prediction model for predicting field strength parameters based on geometric information in the three-dimensional digital model; optimizing the prediction model by using the true value data set, and obtaining an optimized target prediction model; and performing field strength prediction on a to-be-queried point based on the target prediction model, and outputting a prediction result of city wireless signal field strength. The method solves the problem of how to improve the efficiency and accuracy of field strength prediction in the related art.

Description

Technical Field

[0001] This application relates to the field of wireless network simulation and optimization technology, and in particular to a method and system for predicting urban wireless signal field strength based on 3D scene modeling. Background Technology

[0002] In the field of wireless communication, with the rapid development of 5G, the Internet of Things, and low-altitude networks (such as drone communication), the rapid and accurate prediction of wireless signal coverage in large-scale, refined three-dimensional space has become a key aspect of network planning and optimization. Precise field strength (electric field strength) prediction can guide base station deployment, optimize network performance, and ensure communication quality.

[0003] Currently, existing methods for predicting urban wireless signal field strength typically involve photogrammetric modeling followed by ray tracing algorithms. However, these methods suffer from low prediction efficiency and accuracy.

[0004] Therefore, no effective solution has yet been proposed for improving the efficiency and accuracy of existing electric field strength prediction methods. Summary of the Invention

[0005] This application provides a method and system for predicting urban wireless signal field strength based on three-dimensional scene modeling, in order to at least solve the problem of how to improve the efficiency and accuracy of field strength prediction in related technologies.

[0006] In a first aspect, embodiments of this application provide a method for predicting urban wireless signal strength based on three-dimensional scene modeling, the method comprising: Based on urban geographic information data, a three-dimensional digital model is constructed, and parameters are configured on the three-dimensional digital model to generate a wireless signal simulation model; Based on the wireless signal simulation model, the field strength is calculated at preset sampling points to generate a ground truth dataset for training. Based on the geometric information in the three-dimensional digital model, a prediction model for predicting field strength parameters is constructed. The prediction model is optimized using the true value dataset to obtain the optimized target prediction model; Based on the target prediction model, the field strength of the query point is predicted, and the prediction result of the urban wireless signal field strength is output.

[0007] In one embodiment, constructing a three-dimensional digital model based on urban geographic information data includes: Urban geographic information data is acquired, and invalid values ​​are processed from the urban geographic information data to obtain processed geographic information data; the geographic information data includes digital surface models, topographic classification maps, and satellite orthophotos; Based on the terrain classification map, the pixel regions of building categories are extracted to generate Boolean-type building masks; Based on the digital surface model, elevation change regions are extracted using an edge detection algorithm to generate a DSM edge map; the DSM edge map is then ANDed with the building mask to obtain a building edge map. By performing morphological closing operations on the building edge map, broken edge connection and hole filling operations are performed to generate a binarized building footprint map; Based on the building footprint map, a building label map is obtained, and a three-dimensional digital model is generated based on the label map, the digital surface model, and the corresponding digital elevation model.

[0008] In one embodiment, the step of obtaining a building label map based on the building footprint map, and generating a three-dimensional digital model based on the label map, the digital surface model, and the corresponding digital elevation model, includes: The architectural footprint map is subjected to morphological dilation processing to obtain an optimized architectural footprint map; Based on the optimized building footprint map, a unique integer identifier is assigned to each spatially independent building through connectivity component analysis, generating a building label map. Based on the building label map, the digital surface model, and the corresponding digital elevation model, a three-dimensional digital model is generated through height stretching.

[0009] In one embodiment, configuring parameters for the three-dimensional digital model to generate a wireless signal simulation model includes: Based on the electromagnetic wave propagation characteristics of typical urban features, a standard electromagnetic parameter library is pre-set. The standard electromagnetic parameter library contains electromagnetic parameters corresponding to various materials of typical urban features. The electromagnetic parameters include dielectric constant and conductivity. Based on the topographic classification map of the city corresponding to the three-dimensional digital model and the height data of each building in the three-dimensional digital model, determine the material type corresponding to each feature unit in the three-dimensional digital model; Based on the material type corresponding to each land cover unit, the corresponding electromagnetic parameters are matched and retrieved from the standard electromagnetic parameter library, and the corresponding electromagnetic parameters are configured for each land cover unit in the three-dimensional digital model to generate a wireless signal simulation model that includes three-dimensional geometric structure information and electromagnetic response characteristics of land cover.

[0010] In one embodiment, the step of calculating the field strength at preset sampling points based on the wireless signal simulation model to generate a ground truth dataset for training includes: A predetermined number of sampling points are set up in the three-dimensional space of the city, and the three-dimensional spatial coordinates corresponding to each sampling point are determined. Based on the wireless signal simulation model and the three-dimensional spatial coordinates of each sampling point, the reflection, diffraction and penetration propagation process of electromagnetic waves in the three-dimensional space of the target city is simulated by the ray tracing algorithm, and the true value of the field strength corresponding to each sampling point is calculated. Based on the stated ground values ​​of the field strength, a ground value dataset is constructed for training the prediction model.

[0011] In one embodiment, constructing a prediction model for predicting field strength parameters based on the geometric information in the three-dimensional digital model includes: Based on the geometric information in the three-dimensional digital model, multi-view sampling processing is performed on the three-dimensional digital model to obtain scene sampling data; Using the scene sampling data, multiple 3D Gaussian spheres covering three-dimensional space are fitted and generated. The geometric parameters corresponding to each 3D Gaussian sphere are determined, and the field strength parameters are initialized and configured for each 3D Gaussian sphere. The geometric parameters are used to characterize the position, shape, and transparency of the 3D Gaussian sphere in three-dimensional space, and the field strength parameters are used to characterize the contribution of the 3D Gaussian sphere to the spatial distribution of the wireless signal field strength. Based on the geometric parameters and field strength parameters corresponding to the 3D Gaussian sphere, a prediction model for field strength prediction is constructed.

[0012] In one embodiment, optimizing the prediction model using the truth dataset to obtain an optimized target prediction model includes: The field strength parameters in the prediction model are used as the field strength parameters to be optimized, and the geometric parameters in the prediction model are frozen during the training process. The objective is to minimize the deviation between the true field strength values ​​and the predicted field strength values ​​in the true field strength dataset, and to perform iterative optimization on the field strength parameters to be optimized in the prediction model. Based on the preset training convergence condition, the iterative optimization process is terminated, and the optimized target prediction model is obtained.

[0013] In one embodiment, the step of minimizing the deviation between the true field strength values ​​and the predicted field strength values ​​in the true field strength dataset, and performing iterative optimization processing on the field strength parameters to be optimized in the prediction model, includes: Based on the three-dimensional spatial coordinates of each sampling point in the true dataset, the prediction model is used to perform field strength rendering calculations to obtain the predicted field strength value corresponding to each sampling point. Calculate the deviation between the predicted field strength value and the true field strength value in the true value dataset, and construct a loss function with minimizing the deviation as the optimization objective. The field strength parameters to be optimized in the prediction model are iteratively updated by a preset optimizer in order to minimize the loss function.

[0014] In one embodiment, the step of predicting the field strength of the query point based on the target prediction model and outputting the predicted result of the urban wireless signal field strength includes: Based on the wireless signal strength prediction requirements of urban areas, determine the three-dimensional spatial coordinates of one or more query points; Based on the three-dimensional spatial coordinates of the point to be queried and the target prediction model, the predicted field strength value corresponding to the point to be queried is obtained through field strength rendering calculation. Based on the predicted field strength value, the predicted field strength of the urban wireless signal for the corresponding query point is output.

[0015] Secondly, embodiments of this application provide an urban wireless signal strength prediction system based on three-dimensional scene modeling. The system is used to execute the above-described method and includes a model construction module, a ground truth dataset module, a prediction model module, a target prediction model, and an output module; wherein: The model building module is used to build a three-dimensional digital model based on urban geographic information data, configure parameters for the three-dimensional digital model, and generate a wireless signal simulation model. The truth dataset module is used to calculate the field strength of preset sampling points based on the wireless signal simulation model, and generate a truth dataset for training. The prediction model module is used to construct a prediction model for predicting field strength parameters based on the geometric information in the three-dimensional digital model. The target prediction model is used to optimize the prediction model using the true value dataset to obtain an optimized target prediction model. The output module is used to predict the field strength of the query point based on the target prediction model and output the prediction result of the urban wireless signal field strength.

[0016] Thirdly, embodiments of this application provide a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements a method for predicting urban wireless signal field strength based on three-dimensional scene modeling as described in the first aspect above.

[0017] Fourthly, embodiments of this application provide a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements a method for predicting urban wireless signal field strength based on three-dimensional scene modeling as described in the first aspect above.

[0018] The urban wireless signal field strength prediction method and system based on three-dimensional scene modeling provided in this application embodiment has at least the following technical effects.

[0019] Based on urban geographic information data, a 3D digital model is constructed, and parameters are configured on the 3D digital model to generate a wireless signal simulation model. By using readily available urban geographic information data to complete the 3D digital model construction, the cost and data processing complexity of 3D modeling of urban scenes are significantly reduced, solving the problems of cumbersome modeling processes and high computational costs in existing technologies. At the same time, by configuring matching electromagnetic parameters for the 3D digital model, a simulation model with both geometric structure and electromagnetic response characteristics is generated, making up for the lack of key electromagnetic parameters in existing technology models and providing a reliable scene foundation for subsequent high-precision field strength simulation.

[0020] Based on a wireless signal simulation model, the field strength of preset sampling points is calculated to generate a ground truth dataset for training. The field strength of the sampling points is calculated using a wireless signal simulation model with complete electromagnetic parameters, obtaining high-precision ground truth data of field strength. A standardized training dataset is constructed, providing accurate and reliable benchmark labels for the training of subsequent prediction models.

[0021] Based on the geometric information in the three-dimensional digital model, a prediction model for predicting field strength parameters is constructed. The field strength prediction model is built by using the real geometric information of the three-dimensional digital model of the city, so that the prediction model can be adapted to the spatial structure of the urban scene. This solves the problem that the traditional channel model has poor adaptability to complex urban scenes and needs to be switched on demand, laying the model foundation for the rapid prediction of field strength in the future.

[0022] By using the ground truth dataset, the prediction model is optimized to obtain the optimized target prediction model. The prediction model is then iteratively optimized using a high-precision ground truth dataset, enabling the prediction model to have a high-precision prediction capability for the three-dimensional spatial field strength of the city.

[0023] Based on the target prediction model, the field strength of the query point is predicted, and the predicted result of the urban wireless signal field strength is output. The optimized target prediction model completes the field strength prediction for any query point, achieving low latency and fast inference of wireless signal field strength in urban three-dimensional space. While ensuring prediction accuracy, it significantly reduces the computational load of field strength prediction in large-scale urban scenarios, solving the problems of computational explosion, low simulation efficiency, and low prediction accuracy of traditional ray tracing algorithms in complex urban scenarios. It achieves a balance between field strength prediction accuracy, efficiency, and cost, effectively supporting the planning and optimization of urban wireless networks.

[0024] Details of one or more embodiments of this application are set forth in the following drawings and description to make other features, objects and advantages of this application more readily apparent. Attached Figure Description

[0025] The accompanying drawings, which are included to provide a further understanding of this application and form part of this application, illustrate exemplary embodiments and are used to explain this application, but do not constitute an undue limitation of this application. In the drawings: Figure 1 This is a flowchart of a method for predicting urban wireless signal strength based on 3D scene modeling; Figure 2 This is a flowchart illustrating step S101 of constructing a three-dimensional digital model according to an exemplary embodiment; Figure 3 This is a flowchart illustrating step S103 according to an exemplary embodiment; Figure 4 This is a flowchart illustrating step S104 according to an exemplary embodiment; Figure 5 This is a system structure block diagram illustrating an urban wireless signal field strength prediction system based on three-dimensional scene modeling, according to an exemplary embodiment. Figure 6 This is a structural block diagram of an electronic device according to an exemplary embodiment. Detailed Implementation

[0026] To make the objectives, technical solutions, and advantages of this application clearer, the application is described and illustrated below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application. All other embodiments obtained by those skilled in the art based on the embodiments provided in this application without inventive effort are within the scope of protection of this application.

[0027] Obviously, the accompanying drawings described below are merely some examples or embodiments of this application. Those skilled in the art can apply this application to other similar scenarios based on these drawings without any inventive effort. Furthermore, it is understood that although the efforts made in this development process may be complex and lengthy, for those skilled in the art related to the content disclosed in this application, any changes to design, manufacturing, or production based on the technical content disclosed in this application are merely conventional technical means and should not be construed as insufficient disclosure of the content of this application.

[0028] In this application, the reference to "embodiment" means that a specific feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of this phrase in various places in the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment that is mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described in this application may be combined with other embodiments without conflict.

[0029] Unless otherwise defined, the technical or scientific terms used in this application shall have the ordinary meaning understood by one of ordinary skill in the art to which this application pertains. The terms “a,” “an,” “an,” “the,” and similar words used in this application do not indicate quantity limitation and may indicate singular or plural. The terms “comprising,” “including,” “having,” and any variations thereof used in this application are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or device that includes a series of steps or modules (units) is not limited to the listed steps or units, but may also include steps or units not listed, or may include other steps or units inherent to these processes, methods, products, or devices. The terms “connected,” “linked,” “coupled,” and similar words used in this application are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. “Multiple” used in this application refers to two or more. “And / or” describes the relationship between related objects, indicating that three relationships may exist; for example, “A and / or B” can represent: A alone, A and B simultaneously, and B alone. The character " / " generally indicates that the preceding and following objects are in an "or" relationship. The terms "first," "second," and "third" used in this application are merely to distinguish similar objects and do not represent a specific ordering of the objects.

[0030] In the field of wireless communication, with the rapid development of 5G, the Internet of Things, and low-altitude networks (such as drone communication), the rapid and accurate prediction of wireless signal coverage in large-scale, refined three-dimensional space has become a key aspect of network planning and optimization. Accurate field strength prediction can guide base station deployment, optimize network performance, and ensure communication quality. Currently, the technical path for achieving three-dimensional urban wireless signal coverage prediction typically includes two core steps: three-dimensional scene modeling and field strength prediction algorithm simulation.

[0031] In terms of 3D scene modeling, existing technologies are mainly based on photogrammetric modeling, which uses drone aerial images or street view images to generate high-precision 3D models using photogrammetric techniques. While this method can obtain detailed geometric structures, the data processing flow is complex, the computational cost is high, and it usually lacks key electromagnetic parameter information such as the surface material of buildings.

[0032] In terms of field strength prediction algorithms, the traditional ray tracing algorithm is currently recognized as a high-precision prediction method. This algorithm is based on geometric optics and uniform diffraction theory, predicting signal coverage by simulating the propagation paths of electromagnetic waves in three-dimensional space, including reflection, diffraction, and penetration. However, its computational complexity is high. When dealing with large-scale, detailed urban 3D scenes, the number of rays to be traced increases explosively, resulting in a huge computational burden. Furthermore, during ray tracing, the signal intensities of multiple rays arriving at the same receiving point need to be superimposed. When the signal merging operation is called frequently, it severely slows down the overall simulation speed, leading to low planning efficiency.

[0033] In summary, existing technologies have shortcomings in terms of efficiency in 3D scene modeling and matching of electromagnetic parameters. Furthermore, they suffer from severe performance bottlenecks in the core algorithms for field strength prediction, particularly in the signal merging stage. These limitations make it difficult for current signal coverage prediction methods to balance accuracy, efficiency, and cost.

[0034] Based on the above, this application provides a method and system for predicting urban wireless signal field strength in three-dimensional scene modeling.

[0035] In this document, it should be understood that the terms used may be technical means used to implement part of the present invention or other summary technical terms. For example, the terms may include: The true value dataset is a dataset of the three-dimensional spatial coordinates of sampling points and the corresponding true field strength values, obtained by using a wireless signal simulation model with electromagnetic parameters as input and calculating it through a ray tracing algorithm.

[0036] Digital surface model: A digital elevation model of standardized geographic data, including elevation information of all above-ground objects such as bare land, buildings, vegetation, and bridges, reflecting the true three-dimensional surface morphology of the earth's surface.

[0037] Topographic classification map: A thematic map generated after pixel-level land cover classification of remote sensing imagery / geographic raster data. Each pixel in the map is assigned a unique land cover category label (such as buildings, vegetation, water bodies, roads, etc.) to distinguish different types of land cover.

[0038] Elevation abrupt change areas: After processing the digital surface model using edge detection algorithms, areas with drastic changes in elevation values ​​are extracted. These areas correspond to the boundary between the facade of urban buildings and the ground, and are the core feature basis for generating accurate building outlines.

[0039] Building footprint map: A binary closed contour map generated by performing morphological closing operations and hole filling on the filtered building edge map, which fully presents the planar boundary shape of each building.

[0040] Connectivity component analysis: used to identify and label the closed outline of each spatially independent building in the map, assign a unique integer ID to each spatially independent and unconnected building footprint polygon, generate a building label map, and achieve accurate differentiation and independent modeling of individual buildings in the city.

[0041] 3D Gaussian sphere: A field strength parameter used to characterize the spatial distribution of wireless signal field strength; by covering the three-dimensional space of the target city with multiple 3D Gaussian spheres, scene geometric information and field strength distribution information can be encoded simultaneously, enabling high-precision, low-latency, and rapid prediction of urban wireless signal field strength.

[0042] In a first aspect, embodiments of this application provide a method for predicting urban wireless signal strength based on three-dimensional scene modeling. Figure 1 This is a flowchart of a method for predicting urban wireless signal strength based on 3D scene modeling, such as... Figure 1 As shown, the method includes: Step S101: Based on urban geographic information data, construct a three-dimensional digital model, configure parameters for the three-dimensional digital model, and generate a wireless signal simulation model.

[0043] Step S102: Based on the wireless signal simulation model, calculate the field strength of the preset sampling points to generate a ground truth dataset for training.

[0044] Step S103: Based on the geometric information in the three-dimensional digital model, construct a prediction model for predicting field strength parameters.

[0045] Step S104: Optimize the prediction model using the true value dataset to obtain the optimized target prediction model.

[0046] Step S105: Based on the target prediction model, predict the field strength of the query point and output the prediction result of the urban wireless signal field strength.

[0047] In summary, this application provides a method for predicting urban wireless signal field strength based on 3D scene modeling. A 3D digital model is constructed using urban geographic information data, and parameters are configured on the 3D digital model to generate a wireless signal simulation model. By using readily available urban geographic information data to complete the 3D digital model construction, the cost and data processing complexity of urban scene 3D modeling are significantly reduced, solving the problems of cumbersome modeling processes and high computational costs in existing technologies. Simultaneously, by configuring matching electromagnetic parameters for the 3D digital model, a simulation model with both geometric structure and electromagnetic response characteristics is generated, overcoming the deficiency of existing models lacking key electromagnetic parameters and providing a reliable scene foundation for subsequent high-precision field strength simulation. Based on the wireless signal simulation model, field strength is calculated at preset sampling points to generate a ground truth dataset for training. By completing the field strength calculation at sampling points using a wireless signal simulation model with complete electromagnetic parameters, high-precision ground truth field strength data is obtained, and a standardized training dataset is constructed, providing accurate and reliable benchmark labels for subsequent prediction model training. Based on the geometric information in the 3D digital model, a prediction model for predicting field strength parameters is constructed. The field strength prediction model is built using the real geometric information of the city's 3D digital model, adapting the prediction model to the spatial structure of the urban scene. This solves the problem of poor adaptability of traditional channel models to complex urban scenes and the need for on-demand model switching, laying the model foundation for subsequent rapid field strength prediction. The prediction model is optimized using a ground truth dataset to obtain an optimized target prediction model. Iterative optimization of the prediction model using a high-precision ground truth dataset enables it to achieve high-precision prediction of urban 3D spatial field strength. Based on the target prediction model, field strength prediction is performed on the query point, outputting the predicted result of the urban wireless signal field strength. The optimized target prediction model completes the field strength prediction for any query point, achieving low-latency and fast inference of wireless signal field strength in urban 3D space. While ensuring prediction accuracy, it significantly reduces the computational load of field strength prediction in large-scale urban scenes, solving the problem of computational explosion and low simulation efficiency of traditional ray tracing algorithms in complex urban scenes. It achieves a balance between field strength prediction accuracy, efficiency, and cost, effectively supporting the planning and optimization of urban wireless networks.

[0048] In one embodiment, Figure 2 This is a flowchart illustrating step S101 of constructing a three-dimensional digital model according to an exemplary embodiment, such as... Figure 2 As shown, step S101, constructing a three-dimensional digital model based on urban geographic information data, specifically includes the following steps: Step S1011: Obtain urban geographic information data and perform invalid value processing on the urban geographic information data to obtain processed geographic information data; the geographic information data includes digital surface model, digital elevation model, and topographic classification map obtained after processing high-resolution satellite orthophotos.

[0049] Optionally, geographic information data may be acquired, specifically including high-precision digital surface models. and pre-aligned terrain classification map Invalid values ​​in geographic information data are processed, for example, by marking them as non-numerical, to avoid interfering with subsequent calculations.

[0050] Step S1012: Based on the terrain classification map, extract the pixel regions of building categories and generate a Boolean-type building mask.

[0051] Optionally, from topographic classification maps Extract all regions where pixel values ​​equal the building category. Generate a boolean building mask. .

[0052] Step S1013: Based on the digital surface model, extract the elevation change region through the edge detection algorithm to generate the DSM edge map; perform a logical AND operation between the DSM edge map and the building mask to obtain the building edge map.

[0053] Optionally, for digital surface models The data is processed using the Canny edge detection algorithm to find... Locations where elevations change drastically are identified as building edges. This is achieved by adjusting algorithm parameters, including Gaussian blur coefficients. and high and low dual thresholds ( , Noise suppression is performed to generate a DSM edge map. The DSM edge map obtained in step S1013 The building mask generated in step S1012 Perform a logical AND operation to obtain the building edge map. This operation can be represented as:

[0054] Step S1014: Perform broken edge connection and hole filling operations on the building edge map through morphological closing operation to generate a binarized building footprint map.

[0055] Optionally, for the filtered edge map Perform a morphological closing operation. This operation connects the edge lines detected by the Canny algorithm that may contain minor breaks, forming a closed contour. Then, hole filling is performed on the closed contour to generate a binary architectural footprint. .

[0056] Step S1015: Based on the building footprint map, obtain the building label map, and based on the label map, digital surface model, and corresponding digital elevation model, generate a 3D digital model. Specifically, this includes: The architectural footprint map is subjected to morphological expansion processing to obtain an optimized architectural footprint map. Based on the optimized building footprint map, a unique integer identifier is assigned to each spatially independent building through connectivity component analysis, generating a building label map. Based on building label maps, digital surface models, and corresponding digital elevation models, a three-dimensional digital model is generated through height stretching.

[0057] Optionally, the generated building footprint map Perform a small-radius (e.g., 2-3 pixel units) morphological expansion on the final optimized building footprint map. Perform connectivity component analysis, assigning a unique integer ID to each spatially independent, unconnected building footprint polygon. The final output is a label graph. Each pixel's value represents its corresponding building ID. Based on this tag map... and and the corresponding digital elevation model This can be achieved through the height stretching formula:

[0058] Generate the 3D digital model (3D city building digital model) required for subsequent steps. Output the results of the above algorithm as standard city 3D model data (gpkg format) to obtain the white model of the building.

[0059] Compared with existing technologies (such as using low-altitude drones for scanning or using historical map data provided by map service providers), the three-dimensional digital model in this embodiment can obtain building modeling data that meets the requirements of the subsequent 3D Gaussian splashing algorithm by using building segmentation algorithms and building elevation extraction algorithms, which can obtain satellite imagery and urban elevation data at a lower cost. This achieves a balance between modeling accuracy and modeling cost, resulting in higher economic benefits.

[0060] In one embodiment, step S101 configures parameters for the three-dimensional digital model to generate a wireless signal simulation model. Specifically, this includes the following steps: Step 1: Based on the electromagnetic wave propagation characteristics of typical urban features, a standard electromagnetic parameter library is pre-set. The standard electromagnetic parameter library contains electromagnetic parameters corresponding to various materials of typical urban features, including dielectric constant and conductivity.

[0061] Step 2: Based on the topographic classification map of the city corresponding to the 3D digital model and the height data of each building in the 3D digital model, determine the material type corresponding to each feature unit in the 3D digital model.

[0062] Step 3: Based on the material type corresponding to each land cover unit, match and retrieve the corresponding electromagnetic parameters from the standard electromagnetic parameter library, configure the corresponding electromagnetic parameters for each land cover unit in the 3D digital model, and generate a wireless signal simulation model that includes 3D geometric structure information and electromagnetic response characteristics of land cover.

[0063] Optionally, firstly, a library of electromagnetic parameters corresponding to typical urban terrains is defined, including materials such as concrete, glass curtain walls, vegetation, and water bodies. Then, a terrain classification map is used. Acquire regional terrain information. Based on established rules, set surface parameters for the building white model. For buildings with a floor height greater than 50 meters, use a glass curtain wall texture; for buildings with a floor height less than 50 meters, use a concrete texture; for rivers, use a water texture; for vegetation, use a vegetation texture, etc. Finally, according to the mapping rules, generate a 3D urban building wireless signal simulation model. Based on the three-dimensional digital model of urban buildings, it also includes the electromagnetic response characteristics of each surface.

[0064] By configuring matching electromagnetic parameters for the three-dimensional digital model, a simulation model with both geometric structure and electromagnetic response characteristics was generated, which makes up for the lack of key electromagnetic parameters in existing technology models and provides a reliable scene basis for subsequent high-precision field strength simulation.

[0065] In one embodiment, step S102 involves calculating the field strength at preset sampling points based on a wireless signal simulation model to generate a ground truth dataset for training. Specifically, this includes the following steps: Step S1021: Deploy a preset number of sampling points in the three-dimensional space of the city and determine the three-dimensional spatial coordinates corresponding to each sampling point; Step S1022: Based on the wireless signal simulation model and the three-dimensional spatial coordinates of each sampling point, the reflection, diffraction and penetration propagation process of electromagnetic waves in the three-dimensional space of the target city is simulated by the ray tracing algorithm, and the true value of the field strength corresponding to each sampling point is calculated. Step S1023: Based on the ground values ​​of field strength, construct a ground value dataset for training the prediction model.

[0066] Optionally, to train a 3DGS (3D Gaussian Sputtering) model to predict field strength, it is necessary to obtain real field strength data as ground truth. Therefore, firstly, a 3D urban building wireless signal simulation model is used. As input, the selected sampling points in the scene are processed using traditional ray tracing algorithms. High-precision field strength calculations are performed. Subsequently, the ray tracing algorithm utilizes... It includes electromagnetic parameters and geometric structures to accurately simulate the reflection, diffraction, and penetration of electromagnetic waves, calculating the electromagnetic parameters at each sampling point. True value of the received signal field strength at the location Based on the true value of the field strength Generate a ground truth dataset for training the prediction model. .

[0067] Step S102 completes the field strength calculation of the sampling points using a wireless signal simulation model with complete electromagnetic parameters, obtains high-precision true field strength data, constructs a standardized training dataset, and provides accurate and reliable benchmark labels for the subsequent training of the prediction model.

[0068] In one embodiment, Figure 3 This is a flowchart illustrating step S103 according to an exemplary embodiment, as follows: Figure 3 As shown, step S103 involves constructing a prediction model for predicting field strength parameters based on the geometric information in the three-dimensional digital model. This specifically includes the following steps: Step S1031: Based on the geometric information in the 3D digital model, perform multi-view sampling processing on the 3D digital model to obtain scene sampling data; Step S1032: Using scene sampling data, fit and generate multiple 3D Gaussian spheres covering three-dimensional space, determine the geometric parameters corresponding to each 3D Gaussian sphere, and initialize the field strength parameters for each 3D Gaussian sphere; the geometric parameters are used to characterize the position, shape and transparency of the 3D Gaussian sphere in three-dimensional space, and the field strength parameters are used to characterize the contribution of the 3D Gaussian sphere to the spatial distribution of the wireless signal field strength; Step S1033: Based on the geometric parameters and field strength parameters corresponding to the 3D Gaussian sphere, construct a prediction model for field strength prediction.

[0069] Optionally, the input is first defined as: a 3D digital model of the city buildings from the previous step. This model provides precise geometric information about urban buildings. The initialization process follows: based on... The geometric structure information is obtained by sampling photos from multiple perspectives across the entire scene, and a set of 3D Gaussian spheres is trained based on the sampled photos. Each Gaussian sphere... parameter set Include: ① Position: Three-dimensional coordinates .

[0070] ② Covariance: The covariance matrix representing the shape and size of the Gaussian sphere. .

[0071] ③ Opacity: A scalar quantity representing the transparency of a Gaussian sphere. .

[0072] ④ Field strength parameters: Parameters used to fit wireless signal field strength information. It is represented using spherical harmonic coefficients.

[0073] Based on the geometric parameters and field strength parameters corresponding to the 3D Gaussian sphere, a prediction model for field strength prediction is constructed.

[0074] Step S103 establishes a field strength prediction model using the real geometric information of the city's three-dimensional digital model, making the prediction model compatible with the spatial structure of the urban scene. This solves the problem of poor adaptability of traditional channel models to complex urban scenes and the need to switch models on demand, laying the model foundation for rapid prediction of field strength in the future.

[0075] In one embodiment, Figure 4 This is a flowchart illustrating step S104 according to an exemplary embodiment, as follows: Figure 4 As shown, step S104 involves optimizing the prediction model using the ground truth dataset to obtain the optimized target prediction model. This specifically includes the following steps: Step S1041: Use the field strength parameters in the prediction model as the field strength parameters to be optimized, and freeze the geometric parameters in the prediction model during the training process.

[0076] Step S1042: With the objective of minimizing the deviation between the true field strength values ​​and the predicted field strength values ​​in the ground truth dataset, perform iterative optimization processing on the field strength parameters to be optimized in the prediction model. Specifically, this includes: Based on the three-dimensional spatial coordinates of each sampling point in the true dataset, the predicted field strength value corresponding to each sampling point is obtained by performing field strength rendering calculation through the prediction model. Calculate the numerical deviation between the predicted field strength value and the true field strength value in the dataset, and construct a loss function with minimizing the numerical deviation as the optimization objective. The field strength parameters to be optimized in the prediction model are iteratively updated by a preset optimizer in order to minimize the loss function.

[0077] Step S1043: Terminate the iterative optimization process according to the preset training convergence condition to obtain the optimized target prediction model.

[0078] Alternatively, in order to convert the 3DGS model from geometry rendering to field strength prediction while maintaining its efficiency, the following strategy is adopted: Freeze geometry parameters: Freeze the positions of all Gaussian spheres during training. Covariance and opacity Only optimize the field strength parameters of the Gaussian sphere. .

[0079] Define a field strength prediction function based on 3DGS rendering. For any point in space Its field strength prediction value is obtained by analyzing the field strength parameters of all visible Gaussian spheres. The result is obtained by weighted summation, and the summation process is similar to that of volumetric rendering in 3DGS.

[0080] Field strength parameters use spherical harmonic coefficients To encode the Gaussian sphere Information on the field strength contributed.

[0081] Using mean squared error as the loss function The goal is to minimize the difference between 3DGS predictions and ray tracing ground truth.

[0082] Iterative updates using the Adam optimizer , making the loss function minimize.

[0083] Step S104, by freezing the geometric parameters of the 3D Gaussian sphere and optimizing only the field strength parameters using a lightweight training strategy, significantly reduces the number of parameters to be optimized in model training, thereby significantly reducing training computational complexity and computational power consumption. This effectively solves the industry pain point of traditional ray tracing-based high-precision field strength simulation algorithms experiencing explosive computational load and low simulation efficiency in large-scale urban scenarios. Simultaneously, through the field strength prediction function and mean square error loss function of the 3D Gaussian splash rendering logic, high-precision fitting of the urban three-dimensional spatial field strength distribution characteristics is achieved. This improves the efficiency of model training and field strength inference while ensuring consistency between the prediction results and the true values ​​of high-precision ray tracing simulations, balancing the accuracy and efficiency of field strength prediction. Furthermore, compared to existing technologies (either using ray tracing algorithms for high-complexity, high-precision calculations, or calculating channel statistical parameters for channel prediction), the algorithm proposed in this embodiment reduces algorithm complexity while maintaining accuracy, thus enabling rapid inference of signal field strength. Since urban buildings are modeled, it is not necessary to select different channel modeling methods based on different channel characteristics, allowing for adaptive prediction in various scenarios.

[0084] In one embodiment, step S105 involves predicting the field strength of the query point based on the target prediction model and outputting the predicted result of the urban wireless signal field strength. Specifically, this includes the following steps: Step S1051: Based on the wireless signal strength prediction requirements of the urban area, determine the three-dimensional spatial coordinates corresponding to one or more query points; Step S1052: Based on the three-dimensional spatial coordinates of the point to be queried and the target prediction model, the predicted field strength value corresponding to the point to be queried is obtained through field strength rendering calculation. Step S1053: Based on the predicted field strength value, output the predicted field strength of the urban wireless signal for the corresponding query point.

[0085] Optionally, firstly, based on the wireless signal strength prediction requirements of the target urban area, determine the three-dimensional spatial coordinates corresponding to one or more query points; then, using the three-dimensional spatial coordinates of the query points as input, retrieve the complete set of 3D Gaussian sphere parameters from the target prediction model, and through a predefined field strength prediction function adapted to the 3D Gaussian splash volume rendering logic, perform a weighted summation field strength rendering calculation on the field strength parameters of all visible 3D Gaussian spheres covering the spatial range of the query point to obtain the wireless signal strength prediction value corresponding to the query point; finally, based on the calculated field strength prediction value, output the urban wireless signal strength prediction result corresponding to the query point.

[0086] Specifically, after training, the parameters of the 3DGS-based signal field strength prediction algorithm are the optimized set of field strength parameters for all Gaussian spheres.

[0087] Field strength query: For any new query point in the urban space Utilizing the optimized Fast execution of the 3DGS rendering process (i.e., field strength prediction function) Real-time output of urban wireless signal strength prediction results .

[0088] Step S105 completes the field strength prediction of any query point through the optimized target prediction model, realizing low latency and fast inference of wireless signal field strength in urban three-dimensional space. While ensuring prediction accuracy, it significantly reduces the computational load of field strength prediction in large-scale urban scenarios, solves the problem of computational explosion and low simulation efficiency of traditional ray tracing algorithms in complex urban scenarios, and achieves a balance between field strength prediction accuracy, efficiency and cost, effectively supporting the planning and optimization of urban wireless networks.

[0089] In summary, this application provides a method for predicting urban wireless signal field strength based on 3D scene modeling. A 3D digital model is constructed using urban geographic information data, and parameters are configured on the 3D digital model to generate a wireless signal simulation model. By using readily available urban geographic information data to complete the 3D digital model construction, the cost and data processing complexity of urban scene 3D modeling are significantly reduced, solving the problems of cumbersome modeling processes and high computational costs in existing technologies. Simultaneously, by configuring matching electromagnetic parameters for the 3D digital model, a simulation model with both geometric structure and electromagnetic response characteristics is generated, overcoming the deficiency of existing models lacking key electromagnetic parameters and providing a reliable scene foundation for subsequent high-precision field strength simulation. Based on the wireless signal simulation model, field strength is calculated at preset sampling points to generate a ground truth dataset for training. By completing the field strength calculation at sampling points using a wireless signal simulation model with complete electromagnetic parameters, high-precision ground truth field strength data is obtained, and a standardized training dataset is constructed, providing accurate and reliable benchmark labels for subsequent prediction model training. Based on the geometric information in a 3D digital model, a prediction model for predicting field strength parameters is constructed. By using the real geometric information of the city's 3D digital model, a field strength prediction model is built, adapting the prediction model to the spatial structure of the urban scene. This solves the problem of poor adaptability of traditional channel models to complex urban scenes and the need for on-demand model switching, laying the model foundation for subsequent rapid field strength prediction. The prediction model is optimized using a ground truth dataset to obtain an optimized target prediction model. Iterative optimization of the prediction model using a high-precision ground truth dataset enables it to achieve high-precision prediction capabilities for urban 3D spatial field strength. Based on the target prediction model, field strength prediction is performed at the query point, outputting the predicted results of the urban wireless signal field strength. The optimized target prediction model enables the prediction of field strength at any query point, achieving low latency and fast inference of wireless signal field strength in urban three-dimensional space. While ensuring prediction accuracy, it significantly reduces the computational load of field strength prediction in large-scale urban scenarios. It solves the problems of computational explosion, low simulation efficiency, and low prediction accuracy of traditional ray tracing algorithms in complex urban scenarios, achieving a balance between field strength prediction accuracy, efficiency, and cost, and efficiently supporting the planning and optimization of urban wireless networks.

[0090] Secondly, embodiments of this application provide an urban wireless signal field strength prediction system based on three-dimensional scene modeling. Figure 5 This is a system architecture block diagram illustrating an urban wireless signal strength prediction system based on three-dimensional scene modeling, according to an exemplary embodiment. Figure 5 As shown, the system includes a model building module 510, a ground truth dataset module 520, a prediction model module 530, a target prediction model 540, and an output module 550; wherein: The model building module 510 is used to build a three-dimensional digital model based on urban geographic information data, configure parameters for the three-dimensional digital model, and generate a wireless signal simulation model. The truth dataset module 520 is used to calculate the field strength of preset sampling points based on the wireless signal simulation model and generate a truth dataset for training. Prediction model module 530 is used to construct a prediction model for predicting field strength parameters based on geometric information in a three-dimensional digital model. Target prediction model 540 is used to optimize the prediction model using the true value dataset to obtain the optimized target prediction model; The output module 550 is used to predict the field strength of the query point based on the target prediction model and output the prediction result of the urban wireless signal field strength.

[0091] In summary, this application provides an urban wireless signal strength prediction system based on 3D scene modeling. Through the synergistic effect of the model module 510, the ground truth dataset module 520, the prediction model module 530, the target prediction model 540, and the output module 550, it solves the problems of computational explosion, low simulation efficiency, and low prediction accuracy of traditional ray tracing algorithms in complex urban scenarios. It achieves a balance between accuracy, efficiency, and cost in signal strength prediction, effectively supporting the planning and optimization of urban wireless networks.

[0092] It should be noted that the urban wireless signal strength prediction system based on 3D scene modeling provided in this embodiment is used to implement the above-described embodiments, and details already described will not be repeated. As used above, the terms "module," "unit," "subunit," etc., can refer to combinations of software and / or hardware that perform predetermined functions. Although the apparatus described in the above embodiments is preferably implemented in software, hardware implementation, or a combination of software and hardware, is also possible and contemplated.

[0093] Thirdly, embodiments of this application provide an electronic device, Figure 6 This is a block diagram illustrating an electronic device according to an exemplary embodiment. (e.g.) Figure 6 As shown, the electronic device may include a processor 61 and a memory 62 storing computer program instructions.

[0094] Specifically, the processor 61 may include a central processing unit (CPU), an application specific integrated circuit (ASIC), or one or more integrated circuits that can be configured to implement the embodiments of this application.

[0095] The memory 62 may include a large-capacity storage device for data or instructions. For example, and not limitingly, the memory 62 may include a hard disk drive (HDD), a floppy disk drive, a solid-state drive (SSD), flash memory, an optical disk drive, a magneto-optical disk drive, magnetic tape, or a Universal Serial Bus (USB) drive, or a combination of two or more of these. Where appropriate, the memory 62 may include removable or non-removable (or fixed) media. Where appropriate, the memory 62 may be internal or external to a data processing device. In a particular embodiment, the memory 62 is non-volatile memory. In a particular embodiment, the memory 62 includes read-only memory (ROM) and random access memory (RAM). Where appropriate, the ROM may be a mask-programmed ROM, a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), an electrically alterable read-only memory (EAROM), or flash memory, or a combination of two or more of these. Where appropriate, the RAM can be Static Random-Access Memory (SRAM) or Dynamic Random-Access Memory (DRAM). DRAM can be Fast Page Mode Dynamic Random-Access Memory (FPMDRAM), Extended Data Out Dynamic Random-Access Memory (EDODRAM), Synchronous Dynamic Random-Access Memory (SDRAM), etc.

[0096] The memory 62 can be used to store or cache various data files that need to be processed and / or communicated, as well as possible computer program instructions executed by the processor 61.

[0097] The processor 61 reads and executes computer program instructions stored in the memory 62 to implement any of the urban wireless signal strength prediction methods based on three-dimensional scene modeling in the above embodiments.

[0098] In one embodiment, an urban wireless signal strength prediction device based on 3D scene modeling may further include a communication interface 63 and a bus 60. Wherein, as Figure 6 As shown, the processor 61, memory 62, and communication interface 63 are connected through bus 60 and complete communication with each other.

[0099] The communication interface 63 is used to enable communication between the various modules, devices, units, and / or equipment in the embodiments of this application. The communication interface 63 can also enable data communication with other components such as external devices, image / data acquisition devices, databases, external storage, and image / data processing workstations.

[0100] Bus 60 includes hardware, software, or both, that couples together components of an urban wireless signal strength prediction device based on 3D scene modeling. Bus 60 includes, but is not limited to, at least one of the following: Data Bus, Address Bus, Control Bus, Expansion Bus, and Local Bus. For example, and not as a limitation, bus 60 may include Accelerated GraphicsPort (AGP) or other graphics buses, Extended Industry Standard Architecture (EISA) buses, Front Side Bus (FSB), HyperTransport (HT) interconnects, Industry Standard Architecture (ISA) buses, InfiniBand interconnects, Low Pin Count (LPC) buses, memory buses, Micro Channel Architecture (MCA) buses, Peripheral Component Interconnect (PCI) buses, PCI-Express (PCI-X) buses, Serial Advanced Technology Attachment (SATA) buses, Video Electronics Standards Association Local Bus (VLB) buses, or other suitable buses, or combinations of two or more of these. Where appropriate, bus 60 may include one or more buses. Although specific buses are described and illustrated in the embodiments of this application, this application considers any suitable bus or interconnection.

[0101] Fourthly, embodiments of this application provide a computer-readable storage medium storing a program thereon, which, when executed by a processor, implements the urban wireless signal field strength prediction method based on three-dimensional scene modeling provided in the first aspect.

[0102] The readable storage medium may be more specifically adopted, including but not limited to: portable disk, hard disk, random access memory, read-only memory, erasable programmable read-only memory, optical storage device, magnetic storage device, or any suitable combination thereof.

[0103] In a possible implementation, the present invention can also be implemented as a program product comprising program code, which, when the program product is run on a terminal device, causes the terminal device to perform steps of implementing the urban wireless signal field strength prediction method based on three-dimensional scene modeling provided in the first aspect.

[0104] The program code for executing the present invention can be written in any combination of one or more programming languages. The program code can be executed entirely on the user device, partially on the user device, as a standalone software package, partially on the user device and partially on a remote device, or entirely on a remote device.

[0105] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0106] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the invention patent. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this patent application should be determined by the appended claims.

Claims

1. A method for predicting urban wireless signal strength based on 3D scene modeling, characterized in that, The method includes: Based on urban geographic information data, a three-dimensional digital model is constructed, and parameters are configured on the three-dimensional digital model to generate a wireless signal simulation model; Based on the wireless signal simulation model, the field strength is calculated for preset sampling points to generate a ground truth dataset for training. This includes: deploying a preset number of sampling points in the three-dimensional space of the city and determining the three-dimensional spatial coordinates corresponding to each sampling point; simulating the reflection, diffraction, and penetration propagation process of electromagnetic waves in the three-dimensional space of the target city using a ray tracing algorithm based on the wireless signal simulation model and the three-dimensional spatial coordinates of each sampling point, and calculating the ground truth field strength corresponding to each sampling point; and constructing ground truth data for training the prediction model based on the ground truth field strength values. Based on the geometric information in the three-dimensional digital model, a prediction model for predicting field strength parameters is constructed, including: performing multi-view sampling processing on the three-dimensional digital model according to the geometric information in the three-dimensional digital model to obtain scene sampling data; using the scene sampling data, fitting and generating multiple 3D Gaussian spheres covering three-dimensional space, determining the geometric parameters corresponding to each 3D Gaussian sphere, and simultaneously initializing and configuring field strength parameters for each 3D Gaussian sphere; the geometric parameters are used to characterize the position, shape, and transparency of the 3D Gaussian sphere in three-dimensional space, and the field strength parameters are used to characterize the contribution value of the 3D Gaussian sphere to the spatial distribution of wireless signal field strength; based on the geometric parameters and field strength parameters corresponding to the 3D Gaussian spheres, a prediction model for predicting field strength is constructed. Using the ground truth dataset, the prediction model is optimized to obtain an optimized target prediction model, including: using the field strength parameters in the prediction model as field strength parameters to be optimized, freezing the geometric parameters in the prediction model during training; minimizing the deviation between the ground truth field strength values ​​and the predicted field strength values ​​in the ground truth dataset as the objective, performing iterative optimization processing on the field strength parameters to be optimized in the prediction model; terminating the iterative optimization processing according to a preset training convergence condition, thereby obtaining the optimized target prediction model. Based on the target prediction model, the field strength of the query point is predicted, and the prediction result of the urban wireless signal field strength is output.

2. The method for predicting urban wireless signal strength based on three-dimensional scene modeling according to claim 1, characterized in that, The construction of a three-dimensional digital model based on urban geographic information data includes: Urban geographic information data is acquired, and invalid values ​​are processed from the urban geographic information data to obtain processed geographic information data; the geographic information data includes digital surface models, digital elevation models, and topographic classification maps; Based on the terrain classification map, the pixel regions of building categories are extracted to generate Boolean-type building masks; Based on the digital surface model, elevation change regions are extracted using an edge detection algorithm to generate a DSM edge map; the DSM edge map is then ANDed with the building mask to obtain a building edge map. By performing morphological closing operations on the building edge map, broken edge connection and hole filling operations are performed to generate a binarized building footprint map; Based on the building footprint map, a building label map is obtained, and a three-dimensional digital model is generated based on the label map, the digital surface model, and the corresponding digital elevation model.

3. The method for predicting urban wireless signal strength based on three-dimensional scene modeling according to claim 2, characterized in that, The process of obtaining a building label map based on the building footprint map, and generating a three-dimensional digital model based on the label map, the digital surface model, and the corresponding digital elevation model, includes: The architectural footprint map is subjected to morphological dilation processing to obtain an optimized architectural footprint map; Based on the optimized building footprint map, a unique integer identifier is assigned to each spatially independent building through connectivity component analysis, generating a building label map. Based on the building label map, the digital surface model, and the corresponding digital elevation model, a three-dimensional digital model is generated through height stretching.

4. The method for predicting urban wireless signal strength based on three-dimensional scene modeling according to claim 1, characterized in that, The step of configuring parameters for the three-dimensional digital model to generate a wireless signal simulation model includes: Based on the electromagnetic wave propagation characteristics of typical urban features, a standard electromagnetic parameter library is pre-set. The standard electromagnetic parameter library contains electromagnetic parameters corresponding to various materials of typical urban features. The electromagnetic parameters include dielectric constant and conductivity. Based on the topographic classification map of the city corresponding to the three-dimensional digital model and the height data of each building in the three-dimensional digital model, determine the material type corresponding to each feature unit in the three-dimensional digital model; Based on the material type corresponding to each land cover unit, the corresponding electromagnetic parameters are matched and retrieved from the standard electromagnetic parameter library, and the corresponding electromagnetic parameters are configured for each land cover unit in the three-dimensional digital model to generate a wireless signal simulation model that includes three-dimensional geometric structure information and electromagnetic response characteristics of land cover.

5. The method for predicting urban wireless signal strength based on three-dimensional scene modeling according to claim 1, characterized in that, The step of minimizing the deviation between the true field strength values ​​and the predicted field strength values ​​in the true field strength dataset, and performing iterative optimization processing on the field strength parameters to be optimized in the prediction model, includes: Based on the three-dimensional spatial coordinates of each sampling point in the true dataset, the prediction model is used to perform field strength rendering calculations to obtain the predicted field strength value corresponding to each sampling point. Calculate the deviation between the predicted field strength value and the true field strength value in the true value dataset, and construct a loss function with minimizing the deviation as the optimization objective. The field strength parameters to be optimized in the prediction model are iteratively updated by a preset optimizer in order to minimize the loss function.

6. The method for predicting urban wireless signal strength based on three-dimensional scene modeling according to claim 1, characterized in that, The step of predicting the field strength of the query point based on the target prediction model and outputting the predicted result of the urban wireless signal field strength includes: Based on the wireless signal strength prediction requirements of urban areas, determine the three-dimensional spatial coordinates of one or more query points; Based on the three-dimensional spatial coordinates of the point to be queried and the target prediction model, the predicted field strength value corresponding to the point to be queried is obtained through field strength rendering calculation. Based on the predicted field strength value, the predicted field strength of the urban wireless signal for the corresponding query point is output.

7. A system for predicting urban wireless signal strength based on 3D scene modeling, characterized in that, The system is used to execute the method according to any one of claims 1 to 6, and the system includes a model construction module, a ground truth dataset module, a prediction model module, a target prediction model module, and an output module; wherein: The model building module is used to build a three-dimensional digital model based on urban geographic information data, configure parameters for the three-dimensional digital model, and generate a wireless signal simulation model. The truth dataset module is used to calculate the field strength of preset sampling points based on the wireless signal simulation model, and generate a truth dataset for training. The prediction model module is used to construct a prediction model for predicting field strength parameters based on the geometric information in the three-dimensional digital model. The target prediction model module is used to optimize the prediction model using the true value dataset to obtain an optimized target prediction model. The output module is used to predict the field strength of the query point based on the target prediction model and output the prediction result of the urban wireless signal field strength.

Images