A signal coverage identification method, device, equipment, medium and product

Abstract

The present disclosure relates to the technical field of data processing, and in particular provides a signal coverage identification method, device, equipment, medium and product. The method comprises: determining multi-source geospatial features of a signal detection area, and determining base station signal features of the signal detection area; wherein the multi-source geospatial features are used to indicate spatial structure features of the signal detection area; fusing the multi-source geospatial features and the base station signal features to obtain a fusion feature vector; processing the fusion feature vector through a signal classification model to obtain a signal classification result of the signal detection area; wherein the signal classification result is used to indicate signal coverage quality of the signal detection area.

Description

Technical Field

[0001] This disclosure relates to the field of data processing technology, and in particular to a signal coverage identification method, apparatus, device, medium, and product. Background Technology

[0002] With the development of modern wireless communication technology, especially the widespread deployment of 5G networks, signal coverage quality has become one of the important factors affecting user experience and network performance. However, due to the variable environment and complex building designs, the detection of signal coverage quality faces challenges of low adaptability.

[0003] In related technologies, signal coverage quality in various areas can be detected by simulating base station signal transmission using light waves. However, light waves are easily affected by external weather (e.g., light intensity) or building structure type, and the use of light wave propagation simulation tools requires powerful computing resources, resulting in long computation times and frequent model updates. Therefore, when measuring signal coverage quality in areas with different types of building structures and / or different external environments, it is necessary to modify the light wave parameters according to the type of building structure or the specific state of the external environment. This leads to limited generalization ability for signal coverage quality detection in different indoor environments and low time efficiency. Summary of the Invention

[0004] This disclosure is made in view of the above-mentioned problems. This disclosure provides a signal coverage identification method, apparatus, device, medium, and product.

[0005] According to one aspect of this disclosure, a signal coverage identification method is provided, comprising:

[0006] The method involves determining the multi-source geospatial features of a signal detection area and the base station signal features of the signal detection area; wherein the multi-source geospatial features are used to indicate the spatial structure features of the signal detection area.

[0007] The multi-source geospatial features and the base station signal features are fused to obtain a fused feature vector;

[0008] The fused feature vector is processed by a signal classification model to obtain the signal classification result of the signal detection area; wherein the signal classification result is used to indicate the signal coverage quality of the signal detection area.

[0009] Furthermore, according to another embodiment of one aspect of this disclosure, determining the multi-source geospatial features of the signal detection area includes:

[0010] Determine multi-source data for the signal detection area; wherein the multi-source data is used to indicate geospatial information around the signal detection area that affects base station signal transmission;

[0011] The multi-source data is input into a pre-trained model for feature extraction to obtain the geographical features of the signal detection area; wherein, the pre-trained model is trained by a self-supervised contrastive learning method.

[0012] Furthermore, according to another embodiment of one aspect of this disclosure, the multi-source data includes at least one of the following: structural information inside the building in the signal detection area, structural information outside the building in the signal detection area, and structural information of the surrounding environment of the building in the signal detection area.

[0013] Furthermore, according to another embodiment of one aspect of this disclosure, the method further includes training the signal classification model by:

[0014] Acquire training samples and sample labels; wherein, the training samples include base station signal characteristics of the target detection point in each signal transmission frequency band and the geospatial characteristics of the target detection point, and the sample labels are used to indicate the actual signal coverage quality of the target detection point;

[0015] The initial classification model is trained based on the training samples and sample labels to obtain the signal classification model.

[0016] Furthermore, according to another embodiment of one aspect of this disclosure, obtaining the sample label includes:

[0017] Determine the signal receiving power of the target detection point in each of the signal transmission frequency bands;

[0018] Based on the signal receiving power, determine the coverage quality of at least one candidate signal for the target detection point in each of the signal transmission frequency bands;

[0019] Based on the coverage quality of the at least one candidate signal, the sample label of the target detection point in each of the signal transmission frequency bands is determined.

[0020] Furthermore, according to another embodiment of one aspect of this disclosure, determining the sample label of the target detection point in each of the signal transmission frequency bands based on the coverage quality of the at least one candidate signal includes:

[0021] If the number of first-type signal coverage qualities among at least one candidate signal coverage quality of the signal transmission frequency band is greater than a preset threshold, the first-type signal coverage quality is determined as the sample label of the target detection point for the signal transmission frequency band; wherein, the first-type signal coverage quality is used to indicate that the target detection point has no signal coverage.

[0022] Furthermore, according to another embodiment of one aspect of this disclosure, determining the sample label of the target detection point in each of the signal transmission frequency bands based on the coverage quality of the at least one candidate signal includes:

[0023] If the number of first-type signal coverage quality and second-type signal coverage quality in at least one candidate signal coverage quality of the signal transmission frequency band is greater than a preset number threshold, the second-type signal coverage quality is determined as the sample label of the target detection point for the signal transmission frequency band; wherein, the first-type signal coverage quality is used to indicate that the target detection point has no signal coverage, and the second-type signal coverage quality is used to indicate that the signal coverage strength of the target detection point does not meet the preset strength requirement.

[0024] Furthermore, according to another embodiment of one aspect of this disclosure, determining the sample label of the target detection point in each of the signal transmission frequency bands based on the coverage quality of the at least one candidate signal includes:

[0025] If the number of third-type signal coverage qualities in the actual signal coverage quality of the signal transmission frequency band is greater than or equal to a preset threshold, the third-type signal coverage quality is determined as the sample label of the target detection point for the signal transmission frequency band; wherein, the third-type signal coverage quality is used to indicate that the signal coverage strength of the target detection point meets the preset strength requirement.

[0026] Furthermore, according to another embodiment of one aspect of this disclosure, the fused feature vector is processed by a signal classification model to obtain a signal classification result for the signal detection region, including:

[0027] Deep feature data is obtained by extracting features from the fused feature vector using a multilayer perceptron.

[0028] The deep feature data is processed by a normalized exponential function to obtain the signal classification result of the signal detection region.

[0029] Furthermore, according to another embodiment of one aspect of this disclosure, the fusion of the multi-source geospatial features and the base station signal features to obtain a fused feature vector includes:

[0030] The multi-source geospatial features and the base station signal features are weighted and fused using a cross-attention module to obtain the fused feature vector.

[0031] According to another aspect of this disclosure, a signal coverage identification device is provided, comprising:

[0032] A determination module is used to determine the multi-source geospatial features of the signal detection area and the base station signal features of the signal detection area; wherein the multi-source geospatial features are used to indicate the spatial structure features of the signal detection area;

[0033] The fusion module is used to fuse the multi-source geospatial features and the base station signal features to obtain a fused feature vector;

[0034] The classification module is used to process the fused feature vector through a signal classification model to obtain the signal classification result of the signal detection area; wherein the signal classification result is used to indicate the signal coverage quality of the signal detection area.

[0035] According to another aspect of this disclosure, a computer device is provided, including a memory, a processor, and a computer program stored in the memory, wherein the processor executes the computer program to implement the steps of a method for determining a log template.

[0036] According to another aspect of this disclosure, a computer-readable storage medium is provided having a computer program / instructions stored thereon, which, when executed by a processor, implements the steps of a method for determining a log template.

[0037] According to another aspect of this disclosure, a computer program product is provided, including a computer program / instructions that, when executed by a processor, implement steps of a method for determining a log template.

[0038] As will be described in detail below, a signal coverage identification method, apparatus, device, medium, and product according to embodiments of the present disclosure are disclosed. The technical solution of this disclosure first determines the multi-source geospatial features and base station signal features of the signal detection area, and then determines a fused feature vector based on the geospatial features and base station signal features; subsequently, the fused feature vector is processed by a signal classification model to determine the signal coverage quality of the signal detection area. The above-described method of determining the signal coverage quality of the signal detection area through multi-source geospatial features and base station signal features fully considers the geospatial features of the signal detection area, without considering the influence of the external environment (e.g., light intensity) of the signal detection area. Therefore, it eliminates the need to adjust the detection parameters of the signal coverage quality based on the external environment of the signal detection area, thereby improving the generalization of signal coverage quality detection. Simultaneously, since the use of signal propagation simulation tools in the prior art is omitted, the use of computing resources is reduced, thereby improving the time efficiency of signal coverage quality detection.

[0039] It should be understood that both the foregoing general description and the following detailed description are exemplary and intended to provide further illustration of the claimed technology. Attached Figure Description

[0040] The above and other objects, features, and advantages of this disclosure will become more apparent from the more detailed description of the embodiments thereof in conjunction with the accompanying drawings. The drawings are provided to further illustrate the embodiments of this disclosure and form part of the specification. They are used together with the embodiments of this disclosure to explain the disclosure and do not constitute a limitation thereof. In the drawings, the same reference numerals generally represent the same components or steps.

[0041] Figure 1 This is a flowchart of a signal coverage identification method provided in an embodiment of the present disclosure.

[0042] Figure 2 A detailed flowchart of a signal coverage identification method provided in this embodiment of the disclosure.

[0043] Figure 3 This is a schematic diagram of a signal coverage identification device provided in an embodiment of the present disclosure.

[0044] Figure 4 This is a schematic diagram of an electronic device provided in an embodiment of the present disclosure. Detailed Implementation

[0045] To make the objectives, technical solutions, and advantages of this disclosure more apparent, exemplary embodiments according to this disclosure will now be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are merely some embodiments of this disclosure, and not all embodiments of this disclosure. It should be understood that this disclosure is not limited to the exemplary embodiments described herein.

[0046] It should be noted that similar labels and letters in the following figures indicate similar items. Therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures.

[0047] In this document, the term "and / or" merely describes a relationship, indicating that three relationships can exist. For example, A and / or B can represent three cases: A alone, A and B simultaneously, and B alone. Furthermore, the term "at least one" in this document means any combination of at least two of any one or more elements. For example, including at least one of A, B, and C can mean including any one or more elements selected from the set consisting of A, B, and C.

[0048] Research has shown that with the development of modern wireless communication technology, especially the widespread deployment of 5G networks, signal coverage quality has become a crucial factor affecting user experience and network performance. However, due to varying environments and complex building designs, the testing of signal coverage quality faces challenges of low adaptability.

[0049] In related technologies, signal coverage quality in various areas can be detected by simulating base station signal transmission using light waves. However, light waves are easily affected by external weather (e.g., light intensity) or building structure type, and the use of light wave propagation simulation tools requires powerful computing resources, resulting in long computation times and frequent model updates. Therefore, when measuring signal coverage quality in areas with different types of building structures and / or different external environments, it is necessary to modify the light wave parameters according to the type of building structure or the specific state of the external environment. This leads to limited generalization ability for signal coverage quality detection in different indoor environments and low time efficiency.

[0050] Based on the above research, this disclosure provides a signal coverage identification method. First, it determines the multi-source geospatial features and base station signal features of the signal detection area. Then, it determines a fused feature vector based on the geospatial features and base station signal features. Next, it processes the fused feature vector using a signal classification model to determine the signal coverage quality of the signal detection area. This method of determining the signal coverage quality of the signal detection area through multi-source geospatial features and base station signal features fully considers the geospatial features of the signal detection area, eliminating the need to consider the influence of the external environment (e.g., light intensity) and thus eliminating the need to adjust the detection parameters based on the external environment, thereby improving the generalization of signal coverage quality detection. Simultaneously, by omitting the use of signal propagation simulation tools in existing technologies, it reduces the use of computational resources, thereby improving the time efficiency of signal coverage quality detection.

[0051] To facilitate understanding of this embodiment, a signal coverage identification method disclosed in this disclosure will first be described in detail. The execution subject of the signal coverage identification method provided in this disclosure is generally an electronic device with a certain computing power. In some possible implementations, the signal coverage identification method can be implemented by a processor calling computer-readable instructions stored in memory.

[0052] Reference Figure 1 The diagram shows a flowchart of a signal coverage identification method provided in an embodiment of this disclosure. The method includes steps S101 to S103, wherein:

[0053] S101. Determine the multi-source geospatial features of the signal detection area and the base station signal features of the signal detection area; wherein, the multi-source geospatial features are used to indicate the spatial structure features of the signal detection area.

[0054] In embodiments of this disclosure, the signal detection area can be an area inside a building (i.e., indoors) or an area outside a building (i.e., outdoors).

[0055] Here, when the signal detection area is the interior of a building, multi-source geospatial features are used to indicate the geospatial features surrounding the building. These geospatial features include: geospatial features inside the building, geospatial features outside the building, and geospatial features around the building. When the signal detection area is outside a building, multi-source geospatial features are used to indicate the geospatial features outside the building at multiple locations within the signal detection area.

[0056] Here, base station signal characteristics are used to indicate signal transmission parameters and signal propagation parameters. Among them, signal transmission parameters include at least one of the following: the horizontal distance between the base station antenna and the signal detection area, the vertical distance between the base station antenna and the signal detection area, the horizontal angle between the transmission direction of the base station antenna and the signal detection area, the vertical angle between the transmission direction of the base station beam and the signal detection area, the coverage area of ​​the base station beam, and the application scenario of the base station beam.

[0057] The signal propagation parameters include at least one of the following: the vertical characteristics of the base station's transmitted signal during propagation and the horizontal characteristics of the base station's transmitted signal during propagation. The vertical characteristics of the base station's transmitted signal during propagation can be understood as the height difference between the intersection of the base station's transmitted signal and the signal detection area. The horizontal characteristics of the base station's transmitted signal during propagation can be understood as the horizontal distance between the intersection of the base station's transmitted signal and the signal detection area.

[0058] Here, multi-source geospatial features include structural and layout features surrounding the signal detection area. For example, when the signal detection area is inside a building, multi-source geospatial features include structural and layout features inside, outside, and around the building.

[0059] S102. The multi-source geospatial features and base station signal features are fused to obtain the fused feature vector.

[0060] In the embodiments of this disclosure, firstly, feature fusion can be performed on multi-source geospatial features to obtain a fused multi-source geospatial vector. Then, the fused multi-source geospatial vector and base station signal features can be fused to obtain a fused feature vector.

[0061] Here, feature fusion of multi-source geospatial features can be performed through feature concatenation or weighted summation. For example, when the signal detection area is the interior of a building, feature concatenation can be performed on the structural and layout features of the building's interior, exterior, and surrounding areas to obtain a multi-source geospatial vector.

[0062] S103. The fused feature vector is processed by a signal classification model to obtain the signal classification result of the signal detection area; wherein, the signal classification result is used to indicate the signal coverage quality of the signal detection area.

[0063] In embodiments of this disclosure, the fused feature vector can be input into a signal classification model, and the output of the signal classification model can be determined as the signal classification result. The signal classification model can be a Multi-Layer Perceptron (MLP).

[0064] Here, signal coverage quality includes: no signal coverage, weak signal coverage, and normal signal coverage. Among them, the signal received power (RSRP) of normal signal coverage is greater than that of weak signal coverage, and the signal received power of weak signal coverage is greater than that of no signal coverage.

[0065] In the embodiments of this disclosure, firstly, the multi-source geospatial features of the signal detection area and the base station signal features of the signal detection area are determined; wherein, the multi-source geospatial features are used to indicate the spatial structure features of the signal detection area; secondly, the multi-source geospatial features and the base station signal features are fused to obtain a fused feature vector; finally, the fused feature vector is processed by a signal classification model to obtain the signal classification result of the signal detection area; wherein, the signal classification result is used to indicate the signal coverage quality of the signal detection area.

[0066] In the above implementation, firstly, the multi-source geospatial features and base station signal features of the signal detection area are determined, and then a fused feature vector is determined based on the geospatial features and base station signal features. Afterwards, the fused feature vector is processed using a signal classification model to determine the signal coverage quality of the signal detection area. This method of determining the signal coverage quality of the signal detection area through multi-source geospatial features and base station signal features fully considers the geospatial features of the signal detection area, without considering the influence of the external environment (e.g., light intensity). Therefore, it eliminates the need to adjust the signal coverage quality detection parameters based on the external environment of the signal detection area, thereby improving the generalization of signal coverage quality detection. Simultaneously, by omitting the use of signal propagation simulation tools in existing technologies, the use of computational resources is reduced, thereby improving the time efficiency of signal coverage quality detection.

[0067] In an optional embodiment, the above steps determine the multi-source geospatial features of the signal detection area, specifically including the following steps:

[0068] First, determine the multi-source data of the signal detection area; the multi-source data is used to indicate the geospatial information around the signal detection area that affects the base station signal transmission.

[0069] Then, the multi-source data is input into the pre-trained model for feature extraction to obtain the geographical features of the signal detection area; the pre-trained model is trained by a self-supervised contrastive learning method.

[0070] In embodiments of this disclosure, the multi-source data includes at least one of the following: structural information inside the building in the signal detection area, structural information outside the building in the signal detection area, and structural information of the surrounding environment of the building in the signal detection area.

[0071] For example, the structural information within the signal detection area can be determined through an interior layout diagram. For instance, if the signal detection area is inside a building, the structural information within the signal detection area could include information about the rooms, walls, and other physical obstacles inside the building.

[0072] For example, structural information outside the signal detection area can be determined using a building outline map. For instance, if the signal detection area is inside a building, the structural information outside the signal detection area can be the building's external shape and structure.

[0073] For example, satellite maps can be used to determine the structural information surrounding the signal detection area. For instance, if the signal detection area is inside a building, the structural information surrounding the signal detection area can include information about the building's external environment and factors that affect signal propagation (i.e., surrounding terrain, vegetation, etc.).

[0074] After determining the structural information inside the building, outside the building, and the surrounding environment of the building in the signal detection area, the structural information inside the building, outside the building, and the surrounding environment of the building in the signal detection area can be identified as the multi-source data of the signal detection area.

[0075] Here, feature extraction from multi-source data can be performed using the Vision Transformer (ViT) model. That is, the Vision Transformer is determined as a pre-trained model to extract features from multi-source data.

[0076] The target visual converter model can be determined as follows: the initial visual converter model is trained using the self-supervised contrastive learning method of MoCo V3 to obtain the target visual converter model.

[0077] In practice, training image data can be acquired and designated as positive samples. Then, the training image data can be transformed (e.g., rotated, cropped, or noise-added) to obtain negative samples. Next, the initial visual converter model is trained using a self-supervised contrastive learning method combining the positive and negative samples to obtain the target visual converter model. This method utilizes contrastive learning to pre-train the initial visual converter model based on unlabeled data (i.e., training image data) to obtain the target visual converter model. This target visual converter model can capture the influence of multi-source data on signal propagation, thereby improving the accuracy of subsequent processing.

[0078] In an optional embodiment, the signal classification model is trained using the following method:

[0079] First, training samples and sample labels are obtained; the training samples include the base station signal characteristics of the target detection point in each signal transmission frequency band and the geospatial characteristics of the target detection point, and the sample labels are used to indicate the actual signal coverage quality of the target detection point.

[0080] Then, the initial classification model is trained based on the training samples and sample labels to obtain the signal classification model.

[0081] In embodiments of this disclosure, a training detection region can first be determined. Then, representative measurement points can be selected as target detection points within the training detection region. For example, a target detection point can be determined at preset intervals, so that the target detection points can be evenly distributed within the training detection region.

[0082] After identifying the target detection points, the data characteristics of the multi-source data corresponding to each target detection point can be determined, i.e., the geospatial characteristics of the target detection points. Next, the base station signal characteristics of the target detection points in each signal transmission band can be determined. Then, the base station signal characteristics of the target detection points in each signal transmission band and the geospatial characteristics of the target detection points can be used as training samples. The base station signal characteristics in the training samples can be determined through usage logs reported by the user terminals corresponding to the respective base stations.

[0083] After determining the training samples for the target detection points, the signal received power of each target detection point can be measured. Then, the signal received power of each target detection point is merged and classified to obtain the sample labels for each target detection point.

[0084] This involves performing multiple measurements on each target detection point, which means determining the signal receiving power of multiple signals corresponding to each target detection point, in order to ensure the accuracy of subsequent tags for each target detection point.

[0085] Here, the initial classification model can be a multilayer perceptron using Focal Loss as the loss function. The optimizer for the optimization strategy corresponding to the initial classification model is the Adam optimizer, which can be used to update the model parameters of the initial classification model. For example, the learning rate can be dynamically adjusted using a Cosine Annealing (LR) scheduler.

[0086] During the training of the initial classification model, regularization methods can be used to optimize the training. For example, L2 regularization can be used to prevent overfitting; weight decay can be used to control weight growth; dropout can be used to randomly deactivate some neurons to increase generalization ability; and batch normalization can be used to normalize the hidden layer output and stabilize the training process.

[0087] In an optional embodiment, the above steps for obtaining sample labels specifically include the following steps:

[0088] First, determine the signal receiving power of the target detection point in each signal transmission frequency band;

[0089] Secondly, based on the signal receiving power, the coverage quality of at least one candidate signal at the target detection point in each signal transmission frequency band is determined;

[0090] Finally, based on the coverage quality of at least one candidate signal, the sample label of the target detection point in each signal transmission frequency band is determined.

[0091] In embodiments of this disclosure, the signal transmission frequency band of the signal base station corresponding to the training detection area can be adjusted to determine the signal reception power of the target detection point in each signal transmission frequency band.

[0092] Specifically, at least 20 measurements can be performed on each target detection point for different signal transmission frequency bands. For example, by adjusting the signal transmission frequency band to 4G, the signal reception power of each target detection point in the 4G frequency band can be measured 20 times.

[0093] Here, a target power threshold can be set for each signal transmission frequency band. Then, based on the target power threshold, the coverage quality of at least one candidate signal at the target detection point in each signal transmission frequency band is determined. The target power threshold includes a first power threshold and a second power threshold.

[0094] For example, as shown in Table 1, when the signal transmission frequency band is 4G, -120dBm can be determined as the first power threshold, and -110dBm as the second power threshold. If the target detection point is in the 4G signal transmission frequency band and the signal received power is less than the first power threshold, the candidate signal coverage quality of the target detection point in the 4G frequency band is determined to be no signal coverage. If the target detection point is in the 4G signal transmission frequency band and the signal received power is less than the second power threshold but greater than or equal to the first power threshold, the candidate signal coverage quality of the target detection point in the 4G frequency band is determined to be weak signal coverage. If the target detection point is in the 4G signal transmission frequency band and the signal received power is greater than or equal to the second power threshold, the candidate signal coverage quality of the target detection point in the 4G frequency band is determined to be normal signal coverage.

[0095] When the signal transmission frequency band is 5G, -116dBm can be determined as the first power threshold, and -100dBm as the second power threshold. If the target detection point is in the 5G signal transmission frequency band and the signal received power is less than the first power threshold, the candidate signal coverage quality of the target detection point in the 5G frequency band is determined to be no signal coverage. If the target detection point is in the 5G signal transmission frequency band and the signal received power is less than the second power threshold but greater than or equal to the first power threshold, the candidate signal coverage quality of the target detection point in the 5G frequency band is determined to be weak signal coverage. If the target detection point is in the 5G signal transmission frequency band and the signal received power is greater than or equal to the second power threshold, the candidate signal coverage quality of the target detection point in the 5G frequency band is determined to be normal signal coverage.

[0096] Table 1

[0097]

[0098] Here, the coverage quality of the target detection point in each candidate signal band can be determined. Then, based on the coverage quality of all candidate signals, the sample label of the target detection point in that signal transmission band is determined.

[0099] In an optional embodiment, the above steps determine the sample label of the target detection point in each signal transmission frequency band based on the coverage quality of at least one candidate signal, specifically including the following steps:

[0100] If the number of first-class signal coverage qualities among at least one candidate signal coverage quality of a signal transmission frequency band is greater than a preset threshold, the first-class signal coverage quality is determined as the sample label of the target detection point for that signal transmission frequency band; wherein, the first-class signal coverage quality is used to indicate that the target detection point has no signal coverage.

[0101] In embodiments of this disclosure, the number of first-type signal coverage qualities for each target detection point in each candidate signal coverage quality across any signal transmission frequency band can be determined. Then, if this number is determined to be greater than a preset threshold, the first-type signal coverage quality can be identified as the sample label for that target detection point for that signal transmission frequency band.

[0102] Here, a preset quantity threshold can be determined based on the number of candidate signal coverage qualities corresponding to each signal transmission frequency band. For example, 90% of the number of candidate signal coverage qualities can be used as the preset quantity threshold.

[0103] In an optional embodiment, the above steps determine the sample label of the target detection point in each signal transmission frequency band based on the coverage quality of at least one candidate signal, specifically including the following steps:

[0104] If the number of first-type signal coverage quality and second-type signal coverage quality in at least one candidate signal coverage quality of a signal transmission frequency band is greater than a preset number threshold, the second-type signal coverage quality is determined as the sample label of the target detection point for that signal transmission frequency band; wherein, the first-type signal coverage quality is used to indicate that the target detection point has no signal coverage, and the second-type signal coverage quality is used to indicate that the signal coverage strength of the target detection point does not meet the preset strength requirement.

[0105] In embodiments of this disclosure, the number of first-type signal coverage qualities and second-type signal coverage qualities for each target detection point in each candidate signal coverage quality of any signal transmission frequency band can be determined. Then, if it is determined that this number is greater than a preset number threshold, the second-type signal coverage quality can be determined as the sample label of the target detection point for that signal transmission frequency band.

[0106] Here, the preset strength requirement is that the signal received power is greater than or equal to the first received threshold in the target power threshold, and less than the second received threshold in the target received power.

[0107] In an optional embodiment, the above steps determine the sample label of the target detection point in each signal transmission frequency band based on the coverage quality of at least one candidate signal, specifically including the following steps:

[0108] If the number of third-class signal coverage qualities in the actual signal coverage quality of the signal transmission frequency band is greater than or equal to a preset threshold, the third-class signal coverage quality is determined as the sample label of the target detection point for the signal transmission frequency band; wherein, the third-class signal coverage quality is used to indicate that the signal coverage strength of the target detection point meets the preset strength requirement.

[0109] In embodiments of this disclosure, the number of third-type signal coverage qualities for each target detection point in each candidate signal coverage quality across any signal transmission frequency band can be determined. Then, if this number is determined to be greater than a preset threshold, the third-type signal coverage quality can be identified as the sample label for that target detection point for that signal transmission frequency band.

[0110] Here, the signal coverage strength of the target detection point meets the preset strength requirement, that is, the signal received power is greater than or equal to the second power threshold in the target power threshold.

[0111] In an optional embodiment, the above steps process the fused feature vector using a signal classification model to obtain the signal classification result of the signal detection region, specifically including the following steps:

[0112] First, feature extraction is performed on the fused feature vector using a multilayer perceptron to obtain deep feature data;

[0113] Then, the deep feature data is processed by the normalized exponential function to obtain the signal classification results of the signal detection area.

[0114] In embodiments of this disclosure, the signal classification model includes an input layer, a hidden layer, and an output layer. A fused feature vector can be input to the input layer of the signal classification model. The input layer can then send the fused feature vector to the hidden layer.

[0115] Here, the hidden layer uses a three-layer perceptron to extract features from the fused feature vector, obtaining deeper features. These deeper features are then sent to the output layer.

[0116] Here, the output layer is equipped with a normalized exponential function (Softmax), which can be used to process deep feature data and obtain the signal classification results of the signal detection area.

[0117] In an optional embodiment, the above steps fuse multi-source geospatial features and base station signal features to obtain a fused feature vector, specifically including the following steps:

[0118] The cross-attention module is used to perform weighted fusion of multi-source geospatial features and base station signal features to obtain a fused feature vector.

[0119] In the embodiments of this disclosure, a cross attention module (i.e., the cross attention module described above) can be used to perform weighted fusion of multi-source geospatial features and base station signal features.

[0120] Reference Figure 2The diagram shown is a detailed flowchart of a signal coverage identification method provided in an embodiment of this disclosure, wherein:

[0121] S10. Obtain training samples and sample labels; wherein, the training samples include the base station signal characteristics of the target detection point in each signal transmission frequency band and the geospatial characteristics of the target detection point, and the sample labels are used to indicate the actual signal coverage quality of the target detection point.

[0122] Here, after identifying the target detection points, the data characteristics of the multi-source data corresponding to each target detection point can be determined, i.e., the geospatial characteristics of the target detection points. Then, the base station signal characteristics of the target detection points in each signal transmission band can be determined. Finally, the base station signal characteristics of the target detection points in each signal transmission band and the geospatial characteristics of the target detection points can be used as training samples. The aforementioned base station signal characteristics can be determined through usage logs reported by the user terminals corresponding to the respective base stations.

[0123] Here, the signal receiving power of the target detection point in each signal transmission frequency band is determined; based on the signal receiving power, at least one candidate signal coverage quality of the target detection point in each signal transmission frequency band is determined; based on at least one candidate signal coverage quality, the sample label of the target detection point in each signal transmission frequency band is determined.

[0124] S20. Train the initial classification model based on the training samples and sample labels to obtain the signal classification model.

[0125] S30. The initial visual converter model is trained using the self-supervised contrastive learning method of MoCo V3 to obtain a pre-trained model.

[0126] Here, the training image data can be designated as positive samples. Then, the training image data can be transformed (i.e., rotated, cropped, or noise-added) to obtain negative samples. Next, the initial visual converter model is trained using a self-supervised contrastive learning method combining the positive and negative samples to obtain the target visual converter model (i.e., the pre-trained model). In summary, by using contrastive learning to pre-train the initial visual converter model based on unlabeled data (i.e., the training image data), the target visual converter model is obtained. This allows the target visual converter model to capture the influence of multi-source data on signal propagation, improving the accuracy of subsequent processing.

[0127] S40. Determine the multi-source data of the signal detection area; wherein, the multi-source data is used to indicate the geospatial information around the signal detection area that affects the base station signal transmission.

[0128] Here, multi-source data includes at least one of the following: structural information inside the building in the signal detection area, structural information outside the building in the signal detection area, and structural information of the surrounding environment of the building in the signal detection area.

[0129] Here, the structural information within the signal detection area can be determined through an interior layout diagram. For example, if the signal detection area is inside a building, the structural information within the signal detection area can include information about the rooms, walls, and other physical obstacles inside the building.

[0130] Here, structural information outside the signal detection area can be determined using a building outline map. For example, if the signal detection area is inside a building, the structural information outside the signal detection area can be the building's external shape and structure.

[0131] Here, structural information surrounding the signal detection area can be determined using satellite maps. For example, if the signal detection area is inside a building, the structural information surrounding the signal detection area can include information about the building's external environment and factors affecting signal propagation (i.e., surrounding terrain, vegetation, etc.).

[0132] Here, after determining the structural information inside the building in the signal detection area, the structural information outside the building in the signal detection area, and the structural information of the surrounding environment of the building in the signal detection area, the structural information inside the building in the signal detection area, the structural information outside the building in the signal detection area, and the structural information of the surrounding environment of the building in the signal detection area can be defined as the multi-source data of the signal detection area.

[0133] S50. Input multi-source data into the pre-trained model for feature extraction to obtain the geographical features of the signal detection area.

[0134] S60. The multi-source geospatial features and base station signal features are fused to obtain the fused feature vector.

[0135] Here, the cross-attention module is used to perform weighted fusion of multi-source geospatial features and base station signal features to obtain a fused feature vector.

[0136] S70. The fused feature vector is processed by a signal classification model to obtain the signal classification result of the signal detection area; wherein, the signal classification result is used to indicate the signal coverage quality of the signal detection area.

[0137] Here, the signal classification model is a multi-layer perceptron (MLP).

[0138] Signal coverage quality includes: no signal coverage, weak signal coverage, and normal signal coverage. The signal received power (RSRP) of normal signal coverage in the signal detection area is greater than that of weak signal coverage, which is greater than that of no signal coverage.

[0139] In actual operation, this solution can produce the following effects:

[0140] (1) Signal coverage analysis method combined with pre-trained model: This scheme improves the efficiency of signal coverage analysis by combining a small amount of measurement data with a pre-trained deep learning model.

[0141] (2) Feature fusion technology of multi-source data: Effectively integrate the features of multi-source data such as building outline, satellite map and indoor layout map to improve the accuracy of signal coverage analysis.

[0142] (3) Signal coverage quality identification technology: By combining a multilayer sensor with Softmax, the accuracy of signal coverage quality in the coverage area is achieved.

[0143] Based on the same inventive concept, this disclosure also provides a signal coverage identification device corresponding to the signal coverage identification method. Since the principle of the device in this disclosure is similar to the above-mentioned signal coverage identification method in this disclosure, the implementation of the device can refer to the implementation of the method, and the repeated parts will not be described again.

[0144] Reference Figure 3 The diagram shown is a schematic of a signal coverage identification device provided in an embodiment of this disclosure. The device includes: a determination module 11, a fusion module 12, and a classification module 13; wherein:

[0145] A determination module is used to determine the multi-source geospatial features of the signal detection area and the base station signal features of the signal detection area; wherein the multi-source geospatial features are used to indicate the spatial structure features of the signal detection area;

[0146] The fusion module is used to fuse the multi-source geospatial features and the base station signal features to obtain a fused feature vector;

[0147] The classification module is used to process the fused feature vector through a signal classification model to obtain the signal classification result of the signal detection area; wherein the signal classification result is used to indicate the signal coverage quality of the signal detection area.

[0148] Specifically, the determining module is also used to determine multi-source data of the signal detection area; wherein, the multi-source data is used to indicate the geospatial information around the signal detection area that affects the base station signal transmission;

[0149] The multi-source data is input into a pre-trained model for feature extraction to obtain the geographical features of the signal detection area; wherein, the pre-trained model is trained by a self-supervised contrastive learning method.

[0150] Specifically, the classification module is also used to acquire training samples and sample labels; wherein, the training samples include base station signal characteristics of the target detection point in each signal transmission frequency band and the geospatial characteristics of the target detection point, and the sample labels are used to indicate the actual signal coverage quality of the target detection point;

[0151] The initial classification model is trained based on the training samples and sample labels to obtain the signal classification model.

[0152] Furthermore, the classification module is also used to determine the signal receiving power of the target detection point in each of the signal transmission frequency bands;

[0153] Based on the signal receiving power, determine the coverage quality of at least one candidate signal for the target detection point in each of the signal transmission frequency bands;

[0154] Based on the coverage quality of the at least one candidate signal, the sample label of the target detection point in each of the signal transmission frequency bands is determined.

[0155] Furthermore, the classification module is also used to determine the first type of signal coverage quality as the sample label of the target detection point for the signal transmission frequency band when the number of first type signal coverage quality among at least one candidate signal coverage quality of the signal transmission frequency band is greater than a preset number threshold; wherein, the first type of signal coverage quality is used to indicate that the target detection point has no signal coverage.

[0156] Furthermore, the classification module is also used to determine the second type of signal coverage quality as the sample label of the target detection point for the signal transmission frequency band when the number of the first type of signal coverage quality and the second type of signal coverage quality in at least one candidate signal coverage quality of the signal transmission frequency band is greater than a preset number threshold; wherein, the first type of signal coverage quality is used to indicate that the target detection point has no signal coverage, and the second type of signal coverage quality is used to indicate that the signal coverage strength of the target detection point does not meet the preset strength requirement.

[0157] Furthermore, the classification module is also used to determine the third type of signal coverage quality as the sample label of the target detection point for the signal transmission frequency band when the number of third type signal coverage quality in the actual signal coverage quality of the signal transmission frequency band is greater than or equal to a preset number threshold; wherein, the third type of signal coverage quality is used to indicate that the signal coverage strength of the target detection point meets the preset strength requirement.

[0158] Furthermore, the classification module is also used to extract features from the fused feature vector using a multilayer perceptron to obtain deep feature data;

[0159] The deep feature data is processed by a normalized exponential function to obtain the signal classification result of the signal detection region.

[0160] Specifically, the fusion module is also used to perform weighted fusion of the multi-source geospatial features and the base station signal features through the cross-attention module to obtain the fused feature vector.

[0161] This embodiment first determines the multi-source geospatial features and base station signal features of the signal detection area, and then determines a fused feature vector based on these features. Next, the fused feature vector is processed using a signal classification model to determine the signal coverage quality of the signal detection area. This method of determining the signal coverage quality of the signal detection area using multi-source geospatial features and base station signal features fully considers the geospatial features of the signal detection area, eliminating the need to consider the influence of the external environment (e.g., light intensity). Therefore, it eliminates the need to adjust the signal coverage quality detection parameters based on the external environment, thereby improving the generalization of signal coverage quality detection. Simultaneously, by omitting the use of signal propagation simulation tools found in existing technologies, it reduces the use of computational resources, thus improving the time efficiency of signal coverage quality detection.

[0162] The processing flow of each module in the device and the interaction flow between each module can be referred to the relevant descriptions in the above method embodiments, and will not be detailed here.

[0163] Corresponding to Figure 1 In addition to the text category detection method in this disclosure, this embodiment also provides an electronic device 400, such as... Figure 4 The diagram shown is a structural schematic of an electronic device 400 provided in an embodiment of this disclosure, including:

[0164] The system includes a processor 41, a memory 42, and a bus 43. The memory 42 stores execution instructions and includes main memory 421 and external memory 422. The main memory 421, also called internal memory, temporarily stores the computational data in the processor 41, as well as data exchanged with external memory such as a hard disk. The processor 41 exchanges data with the external memory 422 through the main memory 421. When the electronic device 400 is running, the processor 41 communicates with the memory 42 through the bus 43, causing the processor 41 to execute the following instructions:

[0165] The method involves determining the multi-source geospatial features of a signal detection area and the base station signal features of the signal detection area; wherein the multi-source geospatial features are used to indicate the spatial structure features of the signal detection area.

[0166] The multi-source geospatial features and the base station signal features are fused to obtain a fused feature vector;

[0167] The fused feature vector is processed by a signal classification model to obtain the signal classification result of the signal detection area; wherein the signal classification result is used to indicate the signal coverage quality of the signal detection area.

[0168] The basic principles of this disclosure have been described above with reference to specific embodiments. However, it should be noted that the advantages, benefits, and effects mentioned in this disclosure are merely examples and not limitations, and should not be considered as essential features of each embodiment of this disclosure. Furthermore, the specific details disclosed above are for illustrative and facilitative purposes only, and are not limitations. These details do not limit the scope of this disclosure to the necessity of employing the aforementioned specific details for implementation.

[0169] The block diagrams of devices, apparatuses, devices, and systems disclosed herein are merely illustrative examples and are not intended to require or imply that they must be connected, arranged, or configured in the manner shown in the block diagrams. As those skilled in the art will recognize, these devices, apparatuses, devices, and systems can be connected, arranged, and configured in any manner. Words such as “comprising,” “including,” “having,” etc., are open-ended terms meaning “including but not limited to,” and are used interchangeably with them. The terms “or” and “and” as used herein refer to the terms “and / or,” and are used interchangeably with them unless the context clearly indicates otherwise. The term “such as” as used herein refers to the phrase “such as but not limited to,” and is used interchangeably with it.

[0170] Additionally, as used herein, the “or” used in a list of items beginning with “at least one” indicates a separate list, such that a list of, for example, “at least one of A, B, or C” means A or B or C, or AB or AC or BC, or ABC (i.e., A and B and C). Furthermore, the word “exemplary” does not imply that the described example is preferred or better than other examples.

[0171] It should also be noted that in the systems and methods of this disclosure, the components or steps can be decomposed and / or recombined. These decompositions and / or recombinations should be considered as equivalent solutions to this disclosure.

[0172] Various changes, substitutions, and modifications can be made to the technology described herein without departing from the teachings defined by the appended claims. Furthermore, the scope of the claims of this disclosure is not limited to the specific aspects of the processes, machines, manufactures, events, means, methods, and actions described above. Currently existing or later-developed processes, machines, manufactures, events, means, methods, or actions that perform substantially the same function or achieve substantially the same result as the corresponding aspects described herein can be utilized. Therefore, the appended claims include such processes, machines, manufactures, events, means, methods, or actions within their scope.

[0173] The above description of the disclosed aspects is provided to enable any person skilled in the art to make or use this disclosure. Various modifications to these aspects will be readily apparent to those skilled in the art, and the general principles defined herein may be applied to other aspects without departing from the scope of this disclosure. Therefore, this disclosure is not intended to be limited to the aspects shown herein, but rather to be carried out within the widest scope consistent with the principles and novel features disclosed herein.

[0174] The above description has been given for purposes of illustration and description. Furthermore, this description is not intended to limit the embodiments of this disclosure to the forms disclosed herein. Although numerous exemplary aspects and embodiments have been discussed above, those skilled in the art will recognize certain variations, modifications, alterations, additions, and sub-combinations therein.

Claims

1. A signal coverage identification method, characterized in that, include: The method involves determining the multi-source geospatial features of a signal detection area and the base station signal features of the signal detection area; wherein the multi-source geospatial features are used to indicate the spatial structure features of the signal detection area. The multi-source geospatial features and the base station signal features are fused to obtain a fused feature vector; The fused feature vector is processed by a signal classification model to obtain the signal classification result of the signal detection area; wherein the signal classification result is used to indicate the signal coverage quality of the signal detection area.

2. The method as described in claim 1, characterized in that, The multi-source geospatial features for determining the signal detection area include: Determine multi-source data for the signal detection area; wherein the multi-source data is used to indicate geospatial information around the signal detection area that affects base station signal transmission; The multi-source data is input into a pre-trained model for feature extraction to obtain the geographical features of the signal detection area; wherein, the pre-trained model is trained by a self-supervised contrastive learning method.

3. The method as described in claim 2, characterized in that, The multi-source data includes at least one of the following: structural information inside the building in the signal detection area, structural information outside the building in the signal detection area, and structural information of the surrounding environment of the building in the signal detection area.

4. The method as described in claim 1, characterized in that, The method also includes training the signal classification model using the following method: Acquire training samples and sample labels; wherein, the training samples include base station signal characteristics of the target detection point in each signal transmission frequency band and the geospatial characteristics of the target detection point, and the sample labels are used to indicate the actual signal coverage quality of the target detection point; The initial classification model is trained based on the training samples and sample labels to obtain the signal classification model.

5. The method as described in claim 4, characterized in that, The acquisition of sample labels includes: Determine the signal receiving power of the target detection point in each of the signal transmission frequency bands; Based on the signal receiving power, determine the coverage quality of at least one candidate signal for the target detection point in each of the signal transmission frequency bands; Based on the coverage quality of the at least one candidate signal, the sample label of the target detection point in each of the signal transmission frequency bands is determined.

6. The method as described in claim 5, characterized in that, The step of determining the sample label of the target detection point in each of the signal transmission frequency bands based on the coverage quality of the at least one candidate signal includes: If the number of first-type signal coverage qualities among at least one candidate signal coverage quality of the signal transmission frequency band is greater than a preset threshold, the first-type signal coverage quality is determined as the sample label of the target detection point for the signal transmission frequency band; wherein, the first-type signal coverage quality is used to indicate that the target detection point has no signal coverage.

7. The method as described in claim 5, characterized in that, The step of determining the sample label of the target detection point in each of the signal transmission frequency bands based on the coverage quality of the at least one candidate signal includes: If the number of first-type signal coverage quality and second-type signal coverage quality in at least one candidate signal coverage quality of the signal transmission frequency band is greater than a preset number threshold, the second-type signal coverage quality is determined as the sample label of the target detection point for the signal transmission frequency band; wherein, the first-type signal coverage quality is used to indicate that the target detection point has no signal coverage, and the second-type signal coverage quality is used to indicate that the signal coverage strength of the target detection point does not meet the preset strength requirement.

8. The method as described in claim 5, characterized in that, The step of determining the sample label of the target detection point in each of the signal transmission frequency bands based on the coverage quality of the at least one candidate signal includes: If the number of third-type signal coverage qualities in the actual signal coverage quality of the signal transmission frequency band is greater than or equal to a preset threshold, the third-type signal coverage quality is determined as the sample label of the target detection point for the signal transmission frequency band; wherein, the third-type signal coverage quality is used to indicate that the signal coverage strength of the target detection point meets the preset strength requirement.

9. The method as described in claim 1, characterized in that, The step of processing the fused feature vector using a signal classification model to obtain the signal classification result for the signal detection region includes: Deep feature data is obtained by extracting features from the fused feature vector using a multilayer perceptron. The deep feature data is processed by a normalized exponential function to obtain the signal classification result of the signal detection region.

10. The method as described in claim 1, characterized in that, The process of fusing the multi-source geospatial features and the base station signal features to obtain a fused feature vector includes: The multi-source geospatial features and the base station signal features are weighted and fused using a cross-attention module to obtain the fused feature vector.

11. A signal coverage identification device, characterized in that, include: A determination module is used to determine the multi-source geospatial features of the signal detection area and the base station signal features of the signal detection area; wherein the multi-source geospatial features are used to indicate the spatial structure features of the signal detection area; The fusion module is used to fuse the multi-source geospatial features and the base station signal features to obtain a fused feature vector; The classification module is used to process the fused feature vector through a signal classification model to obtain the signal classification result of the signal detection area; wherein the signal classification result is used to indicate the signal coverage quality of the signal detection area.

12. A computer device, comprising a memory, a processor, and a computer program stored in the memory, characterized in that, The processor executes the computer program to implement the steps of the signal coverage identification method according to claims 1 to 10.

13. A computer-readable storage medium having a computer program / instructions stored thereon, characterized in that, When the computer program / instruction is executed by the processor, it implements the steps of the signal coverage identification method according to claims 1 to 10.

14. A computer program product comprising a computer program / instructions, characterized in that, When the computer program / instruction is executed by the processor, it implements the steps of the signal coverage identification method according to claims 1 to 10.

Images