Wireless signal simulation method and device, equipment, storage medium and program product

Abstract

The invention relates to a wireless signal simulation method and device, equipment, a storage medium and a program product, and relates to the technical field of wireless communication. The method comprises the following steps: determining geographic data and base station distribution data of a target area; and based on the geographic data, the base station distribution data and a pre-trained signal prediction model, determining the wireless signal intensity of at least one height at the target location in the target area. One-step end-to-end prediction can be realized, and the calculation overhead is reduced and fixed. Moreover, compared with a method based on a deterministic model, high-precision environment information does not need to be collected, and the cost of environment information collection is reduced.

Description

Technical Field

[0001] The present application relates to the field of wireless communication technology, in particular to the field of wireless signal simulation technology, and specifically to a wireless signal simulation method, device, equipment, storage medium and program product. Background Art

[0002] The background of wireless signal simulation tasks can be traced back to the rapid development of communication technology and the widespread application of wireless networks. With the rise of communication technologies such as mobile Internet, Internet of Things, 5G (5th Generation Mobile Communication Technology) and 6G (6th Generation Mobile Communication Technology), the demand and complexity of wireless networks have increased dramatically. Wireless networks must not only support the simultaneous connection of large-scale users, but also ensure high-speed, stable and secure data transmission, which undoubtedly places extremely high demands on the performance and efficiency of wireless communication systems.

[0003] In a laboratory environment, due to cost and technical limitations, it is impossible to completely replicate the propagation characteristics of wireless signals in the real world, such as multipath loss, shadowing effect, frequency selective fading, etc. Therefore, through wireless simulation technology, researchers can test and optimize various wireless communication protocols, network architectures, and device performance in a computer simulation environment, and even predict the behavior and performance of future networks, so as to conduct risk assessment and performance optimization before actual deployment.

[0004] The purpose of the wireless simulation task is to improve the reliability and efficiency of wireless communication systems through simulation and prediction, reduce the complexity of network deployment and operation, and provide strong support for the continuous evolution and innovation of wireless networks.

[0005] At present, researchers have developed several methods in the field of wireless signal simulation.

[0006] 1. Methods based on statistical models. That is, the wireless signal is explicitly modeled. The model is based on the measurement data of a specific propagation environment (such as urban, suburban, and rural areas). The statistical relationship between the wireless signal and the input parameters (such as transmitter-receiver distance, frequency, building density, etc.) is derived using large-scale measurement data. Statistical models usually do not consider the detailed characteristics of the environment, so the prediction accuracy at a specific location is low.

[0007] 2. Methods based on deterministic models. These models are based on the physical principles of electromagnetic wave propagation and can use detailed environmental information (such as the shape and material properties of buildings) to predict wireless signals. Specifically, the electromagnetic waves emitted by the signal source are explicitly modeled as a large number of rays, the trajectory of each ray is tracked, the attenuation of the electromagnetic wave is calculated according to the propagation distance, and the rays reflected or refracted continue to be tracked when the rays reach the surface of the object. Finally, the distribution of rays in space is used to simulate the distribution of wireless signals. However, this type of method requires the simulation of a large number of rays, and each ray needs to be tracked and calculated for interaction with the environment, so the calculation cost is high; depending on the complexity of the scene, the amount of calculation required for the interaction between the ray and the scene is also different, resulting in uncertain calculation costs. In addition, the precision of environmental information is required to be high. In order to accurately calculate the path of the ray, as well as the scattering, reflection, and refraction of the ray on the surface of the ground and the surface of the object, accurate environmental information is required, such as the type of surface, the material of the building surface, etc. This places high demands on the collection of environmental information and increases the cost of environmental information collection. Summary of the invention

[0008] The present application provides a wireless signal simulation method, apparatus, device, storage medium and program product to realize model-based wireless signal simulation, reduce and fix the computational overhead of wireless signal simulation. The technical solution of the present application is as follows:

[0009] In order to achieve the above objectives, this application adopts the following technical solutions:

[0010] In a first aspect, an embodiment of the present application provides a wireless signal simulation method, the method comprising: determining geographic data and base station distribution data of a target area;

[0011] Based on the geographic data, the base station distribution data and a pre-trained signal prediction model, a wireless signal strength at at least one height at a target location within the target area is determined.

[0012] The technical solution provided by the embodiment of the present application brings at least the following beneficial effects: by utilizing the powerful fitting ability of the model, the correlation between limited environmental information (geographic data and base station distribution data) and the distribution of wireless signals is directly learned, thereby avoiding the loss of accuracy that may be caused by missing or incorrect environmental information, and ensuring the accuracy of the prediction. By achieving one-step end-to-end prediction, the computational overhead is reduced and fixed. Moreover, compared with the method based on the deterministic model, it is not necessary to collect highly accurate environmental information such as surface type and building surface material, thereby reducing the cost of collecting environmental information.

[0013] In combination with the above-mentioned first aspect, as a possible implementation method, the geographic data includes at least one of altitude information, land feature category information and building height information; the base station distribution data includes at least one of location information, antenna angle, antenna height and communication index.

[0014] In combination with the above-mentioned first aspect, as a possible implementation method, the antenna angle includes at least one of an azimuth angle, a downtilt angle, a roll angle, and an additional electrical downtilt angle; the communication indicators include: at least one of antenna gain, signal frequency, transmit feeder line length, receive feeder line length, transmit loss, receive loss, and noise coefficient.

[0015] In combination with the first aspect, as a possible implementation manner, determining the wireless signal strength at at least one height at a target location in the target area based on the geographic data, the base station distribution data and the pre-trained signal prediction model includes:

[0016] Processing the geographic data into a geographic data feature map; wherein the two-dimensional coordinate points of the geographic data feature map correspond to sub-areas in the target area; the geographic data feature map includes: at least one of the ground feature category label data, relative altitude data, and relative building height data of the sub-areas corresponding to each of the two-dimensional coordinate points;

[0017] Processing the base station distribution data into a base station distribution characteristic map; wherein the two-dimensional coordinate points of the base station distribution characteristic map correspond to sub-areas in the target area; the base station distribution characteristic map includes: at least one of antenna height relative data, antenna angle characteristic data and communication index characteristic data of base stations in a target sub-area in the target area; the target sub-area is a sub-area where base stations are distributed;

[0018] Generate a splicing feature map based on the geographic data feature map and the base station distribution feature map;

[0019] The spliced ​​feature map is processed based on a pre-trained signal prediction model to obtain the wireless signal strength.

[0020] In combination with the first aspect above, as a possible implementation manner, the method further includes:

[0021] Modeling is performed based on the geographic data and base station distribution data of the target area to obtain a 3D modeling of the target area;

[0022] In the 3D modeling of the target area, the wireless signal strength at at least one height at the target location in the target area output by the signal prediction model is visualized and rendered.

[0023] A second aspect of the embodiments of the present application provides a method for training a signal prediction model, the method comprising:

[0024] Acquire a training set, the training set including geographic data and base station distribution data of a target area, and a true strength of a wireless signal at at least one height at a target location within the target area;

[0025] Inputting the geographic data and the base station distribution data into a signal prediction model to be trained to obtain a predicted wireless signal strength at at least one height at the target location;

[0026] Based on the difference between the actual strength of the wireless signal and the predicted strength of the wireless signal, the parameters of the signal prediction model are adjusted.

[0027] The technical solution provided by the embodiments of the present application brings at least the following beneficial effects:

[0028] By using the powerful fitting ability of the model, the relationship between limited environmental information (geographic data and base station distribution data) and wireless signal distribution is directly learned, avoiding the loss of accuracy caused by missing or incorrect environmental information, and ensuring the accuracy of the prediction. Based on the one-step end-to-end prediction, the computational overhead is reduced and fixed. In addition, compared with the method based on deterministic models, it does not need to collect highly accurate environmental information such as surface type and building surface materials, which reduces the cost of environmental information collection.

[0029] According to a third aspect of the present application, a wireless signal simulation device is provided, the device comprising:

[0030] A first determination module, used to determine geographic data and base station distribution data of a target area;

[0031] The second determination module is used to determine the wireless signal strength at at least one height at a target location in the target area based on the geographic data, the base station distribution data and a pre-trained signal prediction model.

[0032] In combination with the above-mentioned third aspect, as a possible implementation method, the geographic data includes at least one of altitude information, land feature category information and building height information; the base station distribution data includes at least one of location information, antenna angle, antenna height and communication index.

[0033] In combination with the above-mentioned third aspect, as a possible implementation method, the antenna angle includes at least one of an azimuth angle, a downtilt angle, a roll angle, and an additional electrical downtilt angle; the communication indicators include: at least one of antenna gain, signal frequency, transmit feeder line length, receive feeder line length, transmit loss, receive loss and noise coefficient.

[0034] In combination with the third aspect, as a possible implementation manner, the second determining module is specifically configured to:

[0035] Processing the geographic data into a geographic data feature map; wherein the two-dimensional coordinate points of the geographic data feature map correspond to sub-areas in the target area; the geographic data feature map includes: at least one of the ground feature category label data, relative altitude data, and relative building height data of the sub-areas corresponding to each of the two-dimensional coordinate points;

[0036] Processing the base station distribution data into a base station distribution characteristic map; wherein the two-dimensional coordinate points of the base station distribution characteristic map correspond to sub-areas in the target area; the base station distribution characteristic map includes: at least one of antenna height relative data, antenna angle characteristic data and communication index characteristic data of base stations in a target sub-area in the target area; the target sub-area is a sub-area where base stations are distributed;

[0037] Generate a splicing feature map based on the geographic data feature map and the base station distribution feature map;

[0038] The spliced ​​feature map is processed based on a pre-trained signal prediction model to obtain the wireless signal strength.

[0039] According to a fourth aspect of an embodiment of the present application, a training device for a signal prediction model is provided, the device comprising:

[0040] An acquisition module, configured to acquire a training set, wherein the training set includes geographic data and base station distribution data of a target area, and a true strength of a wireless signal at at least one height at a target location within the target area;

[0041] An input module, used for inputting the geographic data and the base station distribution data into a signal prediction model to be trained, so as to obtain a predicted strength of a wireless signal at at least one height at the target location;

[0042] The adjustment module is used to adjust the parameters of the signal prediction model based on the difference between the actual strength of the wireless signal and the predicted strength of the wireless signal.

[0043] According to a fifth aspect of an embodiment of the present application, an electronic device is provided, comprising: a processor; and a memory for storing processor-executable instructions; wherein the processor is configured to execute instructions to implement the method of the first aspect and any possible implementation manner thereof.

[0044] According to a sixth aspect of an embodiment of the present application, a computer-readable storage medium is provided. When instructions in the computer-readable storage medium are executed by a processor of an electronic device, the electronic device is enabled to execute the method in the above-mentioned first aspect and any possible implementation method thereof.

[0045] According to the seventh aspect provided by the present application, a computer program product is provided, the computer program product comprising computer instructions, and when the computer instructions are executed on an electronic device, the electronic device executes the method of the above-mentioned first aspect and any possible implementation manner thereof.

[0046] It should be noted that the technical effects brought about by any implementation method in the second to seventh aspects can refer to the technical effects brought about by the corresponding implementation method in the first aspect, and will not be repeated here. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] Figure 1 A schematic diagram of a flow chart of a wireless simulation method provided in an embodiment of the present application;

[0048] Figure 2 Another flowchart of the wireless simulation method provided in the embodiment of the present application;

[0049] Figure 3 A schematic diagram of data preprocessing provided in an embodiment of the present application;

[0050] Figure 4 A schematic diagram of a flow chart of a method for training a signal strength prediction model provided in an embodiment of the present application;

[0051] Figure 5 A schematic diagram of a model architecture provided in an embodiment of the present application;

[0052] Figure 6 A schematic diagram of the structure of a wireless simulation device provided in an embodiment of the present application;

[0053] Figure 7 A schematic diagram of a structure of a training device for a signal prediction model provided in an embodiment of the present application;

[0054] Figure 8 A schematic diagram of the structure of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0055] The network switching method, device and storage medium according to the embodiments of the present application are described in detail below with reference to the accompanying drawings.

[0056] The term "and / or" in this article is merely a description of the association relationship of associated objects, indicating that three relationships may exist. For example, A and / or B can mean: A exists alone, A and B exist at the same time, and B exists alone.

[0057] The terms "first" and "second" and the like in the specification and drawings of this application are used to distinguish different objects, or to distinguish different processing of the same object, rather than to describe a specific order of objects.

[0058] In addition, the terms "including" and "having" and any variations thereof mentioned in the description of the present application are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device comprising a series of steps or units is not limited to the listed steps or units, but may optionally include other steps or units that are not listed, or may optionally include other steps or units that are inherent to these processes, methods, products or devices.

[0059] It should be noted that, in the embodiments of the present application, words such as "exemplary" or "for example" are used to indicate examples, illustrations or descriptions. Any embodiment or design described as "exemplary" or "for example" in the embodiments of the present application should not be interpreted as being more preferred or more advantageous than other embodiments or designs. Specifically, the use of words such as "exemplary" or "for example" is intended to present related concepts in a specific way.

[0060] In the description of the present application, unless otherwise specified, “plurality” means two or more.

[0061] As introduced in the background technology section, among the existing wireless simulation methods, the statistical model-based method has the technical problem of low prediction accuracy; while the deterministic model-based method needs to simulate a large number of rays in space, which is computationally expensive and uncertain, and it requires a high cost to collect environmental information.

[0062] In response to the above technical problems, the present application provides a wireless signal simulation method, the idea of ​​which is that the environment to be simulated is actually limited, has specific patterns, and has certain rules to follow. The powerful fitting ability of the model is utilized to directly learn the correlation between limited environmental information and wireless signal distribution, thereby achieving one-step end-to-end prediction.

[0063] A wireless simulation method provided in an embodiment of the present application is described in detail below in conjunction with the accompanying drawings of the specification.

[0064] Figure 1 A flowchart of a wireless simulation method provided in an embodiment of the present application, the method comprises the following steps:

[0065] S101: Determine geographic data and base station distribution data of a target area.

[0066] Since the propagation of wireless signals will be affected by the terrain environment, the simulation of wireless signals requires the acquisition of geographical data of the target area, such as the altitude information of the target area, the distribution of buildings, etc. In addition, the distribution of base stations in the target area directly affects the distribution of wireless signals, so it is also necessary to obtain the base station distribution data of the target area, such as the location of the base station in the target area, the antenna height of the base station, etc.

[0067] S102: Determine the wireless signal strength at at least one height at a target location within a target area based on geographic data, base station distribution data, and a pre-trained signal prediction model.

[0068] Specifically, the geographic data and base station distribution data can be preprocessed to obtain feature data that can be directly input into the signal prediction model, and the signal prediction model can be used to obtain the wireless signal strength at at least one height of the target point in the target area. The signal prediction model is pre-trained.

[0069] The target location is the location to be predicted, which can be understood as a target sub-area in the target area. The embodiment of the present application does not limit its area size, nor does it limit the position of the target location in the target area.

[0070] As an example, after determining the target location to be predicted, a square area with a preset area is selected with the target location as the center as the target area, that is, the target location is the center location of the target area.

[0071] The trained signal prediction model can directly output the wireless signal strength at at least one height at the target location, for example, the wireless signal strength at M heights at the target location, where M is a positive integer greater than or equal to 1.

[0072] For the training process of the signal prediction model, please refer to the following.

[0073] The wireless signal simulation method provided in the embodiment of the present application is used to determine the geographic data and base station distribution data of the target area; based on the geographic data, base station distribution data and the pre-trained signal prediction model, the wireless signal strength at at least one height at the target location in the target area is determined. It can be seen that by utilizing the powerful fitting ability of the model, the correlation between limited environmental information (geographic data and base station distribution data) and the distribution of wireless signals is directly learned, which avoids the loss of accuracy that may be caused by the lack or error of environmental information, and can ensure the accuracy of the prediction. By achieving one-step end-to-end prediction, the computational overhead is reduced and fixed. In addition, compared with the method based on the deterministic model, it is not necessary to collect highly accurate environmental information such as surface type and building surface material, which reduces the cost of environmental information collection.

[0074] In one embodiment of the present application, the geographic data includes at least one of altitude information, feature category information, and building height information; the base station distribution data includes at least one of antenna angle, antenna height, and communication index.

[0075] Specifically, the altitude information may be an altitude map, which is used to indicate the altitude of the target area, i.e., the elevation; the feature category information may be a feature category map, which is used to indicate the distribution of various features on the surface of the target area, using different symbols and colors to distinguish and indicate different types of features, such as forests, water bodies, roads, buildings, etc.; the building height information may be a building distribution map, which is used to indicate the distribution of buildings in the target area and the elevation information of the buildings.

[0076] The embodiment of the present application does not limit the format of the above map.

[0077] As an example, the above maps may all be in GIS (Geographic Information System) format, that is, raster data stored row by row with region coordinates and raster resolution.

[0078] One or more of the altitude information, feature category information and building height information can be used to describe the geographical information that affects the transmission and distribution of wireless signals in the target area. Therefore, the above one or more of the information can be used as geographic data for the prediction of wireless signals.

[0079] In one embodiment of the present application, the base station distribution data includes at least one of location information, antenna angle, antenna height and communication index.

[0080] The location information represents the location of the base station in the target area, which may be longitude and latitude information.

[0081] For a base station, the number of sectors can range from one to multiple, depending on the type of base station antenna and coverage requirements. Sector is a geographical concept that represents the coverage area of ​​an antenna. If a base station uses a directional antenna, there can be multiple sectors. For example, each base station can use three directional antennas with 120-degree sector radiation, so each base station has three sectors. Accordingly, when obtaining the antenna angle, antenna height, and communication index, the antenna angle, antenna height, and communication index of each sector of the base station can be obtained.

[0082] In one embodiment of the present application, the antenna angle may include at least one of an azimuth angle, a downtilt angle, a roll angle, and an additional electrical downtilt angle. The communication index may include at least one of an antenna gain, a signal frequency, a transmission feeder length, a reception feeder length, a transmission loss, a reception loss, and a noise factor.

[0083] The above information all affects the propagation and distribution of the signal, so one or more of the above information can be used as base station distribution data for prediction of wireless signals.

[0084] See also Figure 2 In one embodiment of the present application, the step of determining the wireless signal strength at at least one height at a target location within a target area based on geographic data, base station distribution data and a pre-trained signal prediction model specifically includes the following detailed steps:

[0085] S201: Processing geographic data into a geographic data feature map; wherein the two-dimensional coordinate points of the geographic data feature map correspond to sub-areas in the target area; the geographic data feature map includes: at least one of the feature category label data, relative altitude data, and relative building height data of the sub-area corresponding to each two-dimensional coordinate point.

[0086] Specifically, it is necessary to pre-process the geographic data and base station distribution data to obtain feature data that can be received by the model.

[0087] For geographic data, the original data is in the form of a map, that is, it contains information such as the altitude of each location in the target area. In order to simplify the processing, the target area is divided into sub-areas.

[0088] As an example, a rectangular target area of ​​768m×768m may be divided into 256×256 target sub-areas, and the geographical range corresponding to each target sub-area is 3m×3m.

[0089] The geographic data is processed into a three-dimensional geographic data feature map, wherein the two-dimensional coordinate points of the geographic data feature map correspond to a sub-area in the target area, and each two-dimensional coordinate point corresponds to a feature vector, which contains the geographic feature data of the corresponding sub-area.

[0090] Specifically, in order to simplify the model's data processing, the altitude data of the sub-area where the target location is located can be set to a standard value, such as zero. Accordingly, the altitude data and building height data of other sub-areas are adjusted to obtain relative altitude data and relative building height data.

[0091] For a sub-area without a building, the initial value of the building height data may be assigned a specific value, such as zero.

[0092] For feature category information, it can be represented by a one-hot vector, which means using N bits of 0 or 1 to encode N states. For example, the number of feature categories is n, which may include forests, water bodies, roads, buildings, etc., and each component in the corresponding vector takes a value. Then each sub-area corresponds to a specific feature category, and the component corresponding to the geographic category takes a value of 1, and the other n-1 components take a value of 0.

[0093] The finally obtained geographic data feature map contains: geographic data features of the sub-area of ​​each two-dimensional coordinate point, including relative altitude data, relative building height data and feature category label data.

[0094] S202: Process the base station distribution data into a base station distribution feature map; wherein the two-dimensional coordinate points of the base station distribution feature map correspond to sub-areas in the target area; the base station distribution feature map includes: at least one of the antenna height relative data, antenna angle feature data and communication index feature data of the base stations in the target sub-area within the target area; the target sub-area is a sub-area where base stations are distributed.

[0095] Specifically, for base station distribution data, first determine the target sub-area, that is, the sub-area where the base stations are distributed, based on the location information of each base station in the target area. Then, assign the relevant feature data of the base station to the two-dimensional coordinate points corresponding to the sub-area to obtain the base station distribution feature map.

[0096] Among them, the relevant characteristic data of the base station include: relative antenna height data, which is the relative value of the antenna height data relative to the altitude data of the sub-area where the target location is located; the antenna angle characteristic data may include azimuth, downtilt, roll, and additional electrical downtilt, and the sine and cosine values ​​of each angle are calculated separately to form two features; the communication index data may include: antenna gain, signal frequency, transmit feeder line length, receive feeder line length, transmit loss, receive loss, and noise coefficient, and the communication index data can be directly used as features.

[0097] S203: Generate a spliced ​​feature map based on the geographic data feature map and the base station distribution feature map.

[0098] Since both the geographic data feature map and the base station distribution feature map correspond to the same target area, and the sub-areas corresponding to the same two-dimensional coordinate points of the two are the same, the geographic data features and the base station distribution features can be directly spliced ​​to obtain a spliced ​​feature map.

[0099] For easier understanding, further description is given in conjunction with the accompanying drawings. Figure 3 , which is a schematic diagram of data preprocessing provided in an embodiment of the present application, such as Figure 3As shown, the altitude map, building distribution map, feature category map, and base station distribution data are processed separately to obtain corresponding features, and then the features are spliced ​​to obtain a spliced ​​feature map.

[0100] Among them, for the altitude map and building distribution map, it is necessary to perform block cutting and standardization in turn. The block cutting is used to generate sub-regions, so as to establish the correspondence between the two-dimensional coordinate points and the sub-regions in the feature map. For the feature category map, it is necessary to perform block cutting and one-hot vector conversion in turn to obtain the one-hot vector that represents the feature category to which the sub-region belongs. For the base station distribution data, the antenna height, angle and other data are subjected to feature conversion processing to obtain features and assign them to the corresponding two-dimensional coordinate points. Finally, the splicing process is performed to obtain the spliced ​​feature map.

[0101] The final generated spliced ​​feature map covers the relative altitude data, relative building height data, and ground feature category label data of each sub-area in the target area, as well as the relative antenna height data, antenna angle feature data, and communication index feature data of the sub-areas where base stations are distributed.

[0102] S204: Process the spliced ​​feature map based on a pre-trained signal prediction model to obtain wireless signal strength.

[0103] By using the spliced ​​feature map as the input of the signal strength prediction model, the wireless signal strength at at least one height at the target location in the target area output by the model can be obtained.

[0104] It can be seen that the embodiments of the present application provide a data processing technology for processing geographic data and base station distribution data into feature data that can be received by a model, which is conducive to the model learning relevant features to accurately predict the signal strength at a specific location.

[0105] In one embodiment of the present application, in order to intuitively display the geographic information and base station distribution information of the target area, as well as the corresponding wireless signal strength, 3D modeling can be performed based on the geographic data and base station distribution data of the target area to obtain 3D modeling of the target area.

[0106] Specifically, based on the 3D modeling tool, a 3D terrain model can be formed according to the terrain height map, and specific colors can be set for each area of ​​the terrain model according to the terrain category map, and different colors correspond to different terrain types; a 3D building model can be formed according to the building data, and different colors can be randomly set for each building model. From the outside, a corresponding 3D base station model is formed according to the location information of the base station and the angle, height and other information of the antenna and placed in the scene, so as to obtain a 3D modeling of the target area that includes the geographic information of the target area and the base station distribution information.

[0107] In addition, based on the wireless signal strength at at least one height at the target location predicted by the signal prediction model, a signal strength map is drawn in the 3D modeling of the target area, and the strength of the signal can be represented by different colors.

[0108] It can be seen that 3D modeling is performed based on the geographic data and base station distribution data of the target area, and the signal strength predicted by the signal prediction model is visualized in the 3D modeling, which facilitates the intuitive display of the prediction results of the model.

[0109] In one embodiment of the present application, the predicted signal strength and the true signal strength output by the signal prediction model can also be intuitively compared and displayed based on the 3D modeling of the target area. Specifically, for the target area, the true signal strength of at least one height at the target location can be collected, and a signal strength graph can be drawn in the 3D modeling of the target area based on the true signal strength of at least one height at the target location. By switching the true signal strength and the predicted signal strength, it is possible to intuitively display whether the prediction result of the signal prediction model is accurate.

[0110] The following is an introduction to the training process of the signal strength prediction model.

[0111] See also Figure 4 , is a flow chart of a method for training a signal strength prediction model provided in an embodiment of the present application, the method comprising the following steps:

[0112] S401: Acquire a training set, where the training set includes geographic data and base station distribution data of a target area, and actual strength of a wireless signal at at least one height at a target location within the target area.

[0113] Among them, the geographical data and base station distribution data can be found above and will not be repeated here.

[0114] The training set also contains the true strength of the wireless signal at at least one height at the target location in the target area. The target location is the center of the target area.

[0115] For example, the target location is the center of the target area, and the training set includes the actual strength of the wireless signal at the center at an altitude of 100 m, 200 m, and 300 m.

[0116] S402: Inputting geographic data and base station distribution data into a signal prediction model to be trained to obtain a predicted wireless signal strength at at least one height at a target location.

[0117] Specifically, the geographic data and base station distribution data can be preprocessed to obtain the feature map and then input into the signal prediction model to be trained. The preprocessing method can be found above and will not be described here.

[0118] The embodiments of the present application do not limit the model architecture of the signal prediction model. For example, the Transformer architecture, large model architecture (such as ViT, CSWin), deep learning model (such as ResNet, U-Net) can be used, and other machine learning models, such as linear regression models, can also be used.

[0119] The following describes a model architecture that can be used in an embodiment of the present application by way of example.

[0120] The embodiment of the present application may adopt the Swim Transformer model architecture. A sliding window is designed in the Swim Transformer architecture, which calculates self-attention in locally non-overlapping windows and allows cross-window connections.

[0121] See also Figure 5 , is a schematic diagram of a model architecture provided in an embodiment of the present application, such as Figure 5 As shown, the concatenated features (3D feature maps) obtained above are first divided into feature blocks in the spatial dimension, for example, into blocks of size 4×4. Then linear embedding is performed, for example, each block is converted into a C-dimensional embedding vector using a fully connected network. As an example, the value of C can be 48, 96, 192, etc. Thus, they are combined together in the spatial dimension to form a 64×64×C 3D feature map z^0.

[0122] Subsequently, the feature map will be processed by the Transformer architecture. Specifically, it includes 4 sub-networks with similar structures (called "stages"). Stages 1, 2, and 3 contain several sub-networks with the same architecture (called "blocks") and a downsampling layer, and stage 4 contains only several blocks. Among them, in the i-th block, the output z^(i-1) of the previous block is normalized by layer (Layer Normalization, LN) and windowed multi-head self-attention (Windowed Multi-Head SelfAttention, W-MSA), and the residual connection z^(i-1) is added to obtain z~^i, and then z~^i is passed through LN and the fully connected layer, and the residual connection z~^i is added to obtain z^i.

[0123] Among them, W-MSA is based on the Multi-Head Self-Attention (MSA) of the Transformer architecture, with the addition of window restrictions: the input z is divided into several blocks in space with dense non-overlapping 4×4 windows, and the self-attention mechanism is only calculated between the feature vectors in each block, and the self-attention mechanism is not calculated between the feature vectors of different blocks. In addition, a window sliding mechanism is introduced between adjacent blocks: in the W-MSA of the first block in each stage, the origin of the window division of the feature is the upper left corner of the feature. In the W-MSA of the next block, the origin of the window division of the feature is the position deviating from the upper left corner, such as the position of the coordinate (2,2) from the upper left corner. These two window division methods are used alternately in subsequent blocks. In the downsampling layer, the input z is divided into blocks with dense non-overlapping 2×2 windows, and the four vectors in each block are spliced ​​into one vector, and a feature vector with half the dimension is obtained through a fully connected network, and then all the feature vectors are stacked again in space to form a 3D feature map with half the spatial dimension.

[0124] After z^4 is obtained through the fourth stage, it is processed through a pooling layer, that is, it is pooled in two spatial dimensions to form a one-dimensional feature vector, and then the prediction result, that is, the predicted signal value at each height, is obtained through an output network. The output network can be a fully connected network.

[0125] Continuing with the above example, if the set heights are 100m, 200m and 300m, the signal prediction model processes the feature map and outputs the predicted wireless signal strength at 100m, 200m and 300m at the center of the target area.

[0126] S403: Adjust parameters of the signal prediction model based on the difference between the actual strength of the wireless signal and the predicted strength of the wireless signal.

[0127] Specifically, a loss function may be preset, and a loss value may be calculated based on a difference between the actual strength of the wireless signal and the predicted strength of the wireless signal. Parameters in the signal prediction model may be adjusted based on the loss value to implement model training.

[0128] As an example, a mean square error loss function is used to calculate a loss value between the actual strength of the wireless signal and the predicted strength of the wireless signal.

[0129] Through multiple rounds of iterations, the parameters in the signal prediction model are continuously adjusted until the termination condition of the model training is reached, and a trained signal prediction model is obtained. The termination condition of the model training can be set in advance. For example, when the loss function value fluctuates within a certain range and no longer decreases significantly, it can be considered that the model has converged and the training can be terminated.

[0130] It can be seen that the embodiment of the present application provides a method for training a signal prediction model, obtaining a training set, the training set including geographic data and base station distribution data of a target area, and the actual strength of a wireless signal at at least one height at a target location within the target area; inputting the geographic data and the base station distribution data into the signal prediction model to be trained to obtain the predicted strength of the wireless signal at at least one height at the target location; and adjusting the parameters of the signal prediction model based on the difference between the actual strength of the wireless signal and the predicted strength of the wireless signal. The trained signal prediction model can be directly used to predict the signal strength at a specified location within the target area.

[0131] Therefore, by utilizing the powerful fitting ability of the model, the relationship between limited environmental information (geographic data and base station distribution data) and wireless signal distribution can be directly learned, avoiding the loss of accuracy caused by missing or incorrect environmental information, and ensuring the accuracy of the prediction. Based on the one-step end-to-end prediction, the computational overhead is reduced and fixed. In addition, compared with the method based on deterministic models, it does not need to collect highly accurate environmental information such as surface type and building surface materials, which reduces the cost of environmental information collection.

[0132] See also Figure 6 , is a structural diagram of a wireless simulation device provided in an embodiment of the present application, the device comprising:

[0133] A first determination module 601 is used to determine geographic data and base station distribution data of a target area;

[0134] The second determination module 602 is used to determine the wireless signal strength at at least one height at a target location in the target area based on the geographic data, the base station distribution data and a pre-trained signal prediction model.

[0135] Therefore, by utilizing the powerful fitting ability of the model, the relationship between limited environmental information (geographic data and base station distribution data) and wireless signal distribution can be directly learned, avoiding the loss of accuracy caused by missing or incorrect environmental information, and ensuring the accuracy of the prediction. By achieving one-step end-to-end prediction, the computational overhead is reduced and fixed. In addition, compared with the method based on deterministic models, it does not need to collect highly accurate environmental information such as surface type and building surface materials, which reduces the cost of environmental information collection.

[0136] In one possible implementation, the geographic data includes at least one of altitude information, feature category information, and building height information; and the base station distribution data includes at least one of location information, antenna angle, antenna height, and communication index.

[0137] In one possible implementation, the antenna angle includes at least one of an azimuth angle, a downtilt angle, a roll angle, and an additional electrical downtilt angle; the communication indicators include at least one of an antenna gain, a signal frequency, a transmit feeder line length, a receive feeder line length, a transmit loss, a receive loss, and a noise factor.

[0138] In a possible implementation, the second determining module 602 is specifically configured to:

[0139] Processing the geographic data into a geographic data feature map; wherein the two-dimensional coordinate points of the geographic data feature map correspond to sub-areas in the target area; the geographic data feature map includes: at least one of the ground feature category label data, relative altitude data, and relative building height data of the sub-areas corresponding to each of the two-dimensional coordinate points;

[0140] Processing the base station distribution data into a base station distribution characteristic map; wherein the two-dimensional coordinate points of the base station distribution characteristic map correspond to sub-areas in the target area; the base station distribution characteristic map includes: at least one of antenna height relative data, antenna angle characteristic data and communication index characteristic data of base stations in a target sub-area in the target area; the target sub-area is a sub-area where base stations are distributed;

[0141] Generate a splicing feature map based on the geographic data feature map and the base station distribution feature map;

[0142] The spliced ​​feature map is processed based on a pre-trained signal prediction model to obtain the wireless signal strength.

[0143] In a possible implementation, the device further includes:

[0144] A modeling module, used to perform modeling based on the geographic data and base station distribution data of the target area to obtain a 3D modeling of the target area;

[0145] A rendering module is used to perform visual rendering of the wireless signal strength at at least one height at a target location within the target area output by the signal prediction model during the 3D modeling of the target area.

[0146] See also Figure 7 , is a structural schematic diagram of a training device for a signal prediction model provided in an embodiment of the present application, the device comprising:

[0147] An acquisition module 701 is used to acquire a training set, wherein the training set includes geographic data and base station distribution data of a target area, and a true strength of a wireless signal at at least one height at a target location in the target area;

[0148] An input module 702, configured to input the geographic data and the base station distribution data into a signal prediction model to be trained, so as to obtain a predicted wireless signal strength at at least one height at the target location;

[0149] The adjustment module 703 is used to adjust the parameters of the signal prediction model based on the difference between the actual strength of the wireless signal and the predicted strength of the wireless signal.

[0150] In the case of implementing the functions of the above-mentioned integrated modules in the form of hardware, the embodiment of the present application provides a possible structural diagram of the electronic device involved in the above-mentioned embodiment. Figure 8 As shown, the electronic device 800 includes: a processor 802 , a communication interface 803 , and a bus 804 . Optionally, the electronic device 800 may further include a memory 801 .

[0151] The processor 802 is a CPU, a general processor, a network processor (NP), a digital signal processor (DSP), a microprocessor, a microcontroller, a programmable logic device (PLD), or any combination thereof. The processor 802 may also be other devices with processing functions, such as circuits, devices, or software modules, without limitation.

[0152] The bus 804 is used to transmit information between the components included in the electronic device 800 .

[0153] The communication interface 803 is used to communicate with other devices or other communication networks. The other communication networks may be Ethernet, radio access network (RAN), wireless local area network (WLAN), etc. The communication interface 803 may be a module, a circuit, a communication interface or any device capable of achieving communication.

[0154] The memory 801 is used to store instructions, where the instructions may be computer programs.

[0155] Among them, the memory 801 can be a read-only memory (ROM) or other types of static storage devices that can store static information and / or instructions, or a random access memory (RAM) or other types of dynamic storage devices that can store information and / or instructions, or an electrically erasable programmable read-only memory (EEPROM), a compact disc read-only memory (CD-ROM) or other optical disc storage, optical disc storage (including compressed optical disc, laser disc, optical disc, digital versatile disc, Blu-ray disc, etc.), magnetic disk storage medium or other magnetic storage devices, etc., without limitation.

[0156] It should be noted that the memory 801 can exist independently of the processor 802, or can be integrated with the processor 802. The memory 801 can be used to store instructions or program codes or some data, etc. The memory 801 can be located inside the electronic device 800, or outside the electronic device 800, without limitation. The processor 802 is used to execute the instructions stored in the memory 801 to implement the wireless signal simulation method provided in the following embodiments of the present application.

[0157] The present application also provides a computer-readable storage medium having instructions stored thereon. When the instructions in the storage medium are executed by a processor of an electronic device, the electronic device is enabled to execute the wireless signal simulation method provided by the above-mentioned embodiment of the present disclosure.

[0158] The embodiment of the present application also provides a computer program product containing instructions, which, when executed on an electronic device, enables the electronic device to execute the wireless signal simulation method provided by the above-mentioned embodiment of the present disclosure.

[0159] Among them, the computer readable storage medium can be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or device, or any combination of the above. More specific examples of computer readable storage media (a non-exhaustive list) include: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM), a register, a hard disk, an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above, or any other form of computer readable storage medium known in the art. An exemplary storage medium is coupled to a processor so that the processor can read information from the storage medium and write information to the storage medium. Of course, the storage medium can also be a component of the processor. The processor and the storage medium can be located in an application-specific integrated circuit (ASIC). In the embodiments of the present application, a computer-readable storage medium may be any tangible medium that contains or stores a program, which may be used by or in conjunction with an instruction execution system, apparatus, or device.

[0160] The above are only specific implementations of the present application, but the protection scope of the present application is not limited thereto. Any changes or substitutions within the technical scope disclosed in the present application should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. A wireless signal simulation method, characterized in that: The method comprises: Determine the geographic data and base station distribution data of the target area; Based on the geographic data, the base station distribution data and a pre-trained signal prediction model, a wireless signal strength at at least one height at a target location within the target area is determined.

2. The method according to claim 1, characterized in that The geographic data includes at least one of altitude information, feature category information and building height information; the base station distribution data includes at least one of location information, antenna angle, antenna height and communication index.

3. The method according to claim 2, characterized in that The antenna angle includes at least one of azimuth, downtilt, roll, and additional electrical downtilt; the communication indicators include at least one of antenna gain, signal frequency, transmit feeder length, receive feeder length, transmit loss, receive loss, and noise coefficient.

4. The method according to any one of claims 1 to 3, characterized in that: The determining, based on the geographic data, the base station distribution data and the pre-trained signal prediction model, of the wireless signal strength at at least one height at a target location within the target area comprises: Processing the geographic data into a geographic data feature map; wherein the two-dimensional coordinate points of the geographic data feature map correspond to sub-areas in the target area; the geographic data feature map includes: at least one of the ground feature category label data, relative altitude data, and relative building height data of the sub-areas corresponding to each of the two-dimensional coordinate points; Processing the base station distribution data into a base station distribution characteristic map; wherein the two-dimensional coordinate points of the base station distribution characteristic map correspond to sub-areas in the target area; the base station distribution characteristic map includes: at least one of antenna height relative data, antenna angle characteristic data and communication index characteristic data of base stations in a target sub-area in the target area; the target sub-area is a sub-area where base stations are distributed; Generate a splicing feature map based on the geographic data feature map and the base station distribution feature map; The spliced ​​feature map is processed based on a pre-trained signal prediction model to obtain the wireless signal strength.

5. The method according to any one of claims 1 to 3, characterized in that: The method further comprises: Modeling is performed based on the geographic data and base station distribution data of the target area to obtain a 3D modeling of the target area; In the 3D modeling of the target area, the wireless signal strength at at least one height at the target location in the target area output by the signal prediction model is visualized and rendered.

6. A method for training a signal prediction model, characterized in that: The method comprises: Acquire a training set, the training set including geographic data and base station distribution data of a target area, and a true strength of a wireless signal at at least one height at a target location within the target area; Inputting the geographic data and the base station distribution data into a signal prediction model to be trained to obtain a predicted wireless signal strength at at least one height at the target location; Based on the difference between the actual strength of the wireless signal and the predicted strength of the wireless signal, the parameters of the signal prediction model are adjusted.

7. A wireless signal simulation device, characterized in that: The device comprises: A first determination module, used to determine geographic data and base station distribution data of a target area; The second determination module is used to determine the wireless signal strength at at least one height at a target location in the target area based on the geographic data, the base station distribution data and a pre-trained signal prediction model.

8. An electronic device, characterized in that: It comprises a processor and a memory, wherein the processor is coupled to the memory; the memory is used to store computer instructions, and the computer instructions are loaded and executed by the processor to enable the computer device to implement the method according to any one of claims 1 to 5, or the method according to claim 6.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium includes computer-executable instructions, and when the computer-executable instructions are executed on a computer, the computer executes the method according to any one of claims 1 to 5, or the method according to claim 6.

10. A computer program product, characterized in that The computer program product comprises a computer program, and when the computer program is run on an electronic device, the electronic device executes the method according to any one of claims 1 to 5, or the method according to claim 6.

Images