Signal evaluation method, model training method, device, equipment and storage medium

Abstract

The application provides a signal evaluation method and device, a model training method and device, equipment and a storage medium, relates to the technical field of communication, and solves the technical problem that the signal coverage of a region cannot be obtained in time in related technologies. The signal evaluation method comprises the following steps: obtaining device information of an access point device and environment information of the access point device in a first grid, wherein the environment information is used to indicate the degree of influence of signal strength in the first grid; inputting the device information of the access point device and the environment information of the access point device in the first grid into a target signal coverage evaluation model to obtain signal coverage information of the access point device in the first grid.

Description

Technical Field

[0001] This application relates to the field of communication technology, and in particular to a signal evaluation method, model training method, apparatus, device and storage medium. Background Technology

[0002] In evaluating wireless networks, assessing signal coverage is a fundamental and crucial step in network planning. Accurate assessment of wireless network coverage supports the assurance of good wireless network performance metrics, user experience, and resource utilization efficiency.

[0003] Currently, by having surveyors carry measuring equipment to measure the signal strength at multiple locations within an area, the signal coverage of that area can be determined.

[0004] However, the above method, which determines the signal coverage of an area through manual measurement, has the problem of excessive consumption of manpower and material resources, and may not be able to obtain the signal coverage information of the area in a timely manner. Summary of the Invention

[0005] This application provides a signal evaluation method, a model training method, an apparatus, a device, and a storage medium, which solves the technical problem in related technologies that may not be able to obtain the signal coverage of a region in a timely manner.

[0006] In a first aspect, this application provides a signal assessment method, comprising: acquiring device information of an access point device and environmental information of the access point device within a first grid, the environmental information being used to indicate the degree of signal impact within the first grid; inputting the device information of the access point device and the environmental information of the access point device within the first grid into a target signal coverage assessment model to obtain signal coverage information of the access point device within the first grid.

[0007] In this embodiment of the application, by inputting the device information of the access point device and the environmental information of the access point device within the first grid into the target signal coverage evaluation model, the signal coverage within the first grid can be obtained accurately and quickly.

[0008] Optionally, the target signal coverage assessment model is obtained by training an initial signal coverage assessment model based on model training parameters. The model training parameters include: environmental information of the access point device within the first grid, device information of the access point device, first signal coverage information, and second signal coverage information. The environmental information is used to indicate the degree to which the signal strength is affected within the first grid. The first signal coverage information is the actual signal coverage information of the access point device within the first grid, and the second signal coverage information is the actual signal coverage information of the access point device within the second grid. The first grid is smaller than the second grid.

[0009] Secondly, this application provides a signal evaluation device, comprising: an acquisition module and a processing module; the acquisition module is used to acquire device information of an access point device and environmental information of the access point device within a first grid, the environmental information being used to indicate the degree to which signal strength is affected within the first grid; the processing module is used to input the device information of the access point device and the environmental information of the access point device within the first grid into a target signal coverage evaluation model to obtain signal coverage information of the access point device within the first grid.

[0010] Optionally, the target signal coverage assessment model is obtained by training an initial signal coverage assessment model based on model training parameters. The model training parameters include: environmental information of the access point device in the first grid, device information of the access point device, first signal coverage information, and second signal coverage information. The environmental information is used to indicate the degree of signal impact in the first grid. The first signal coverage information is the actual signal coverage information of the access point device in the first grid. The second signal coverage information is the actual signal coverage information of the access point device in the second grid. The first grid is smaller than the second grid.

[0011] Thirdly, this application provides a model training method, comprising: acquiring model training data, the model training data including: environmental information of an access point device within a first grid, device information of the access point device, first signal coverage information, and second signal coverage information, wherein the environmental information is used to indicate the degree to which signal strength is affected within the first grid, the first signal coverage information is the actual signal coverage information of the access point device within the first grid, the second signal coverage information is the actual signal coverage information of the access point device within the second grid, and the first grid is a grid within the second grid; training an initial signal coverage evaluation model based on the model training data to obtain a target signal coverage evaluation model; the target signal coverage evaluation model is used to predict the signal coverage information of the access point device within the first grid based on the environmental information of the access point device within the first grid and the device information of the access point device.

[0012] In this application, since the first grid is a grid within the second grid, the first signal coverage information can more accurately reflect the signal coverage within the first grid than the second signal coverage information, and there is a correlation between the first and second signal coverage information. Because the target signal coverage evaluation model is used to predict the signal coverage information of the access point device within the first grid based on the environmental information and device information of the access point device within the first grid—that is, the first and second signal coverage information serve as the validation set for the initial signal coverage evaluation model during training—the initial signal coverage evaluation model can be trained using both the more accurate first signal coverage information and the less accurate second signal coverage information, resulting in a higher prediction accuracy for the target signal coverage evaluation model.

[0013] Optionally, training the initial signal coverage evaluation model based on the model training data to obtain the target signal coverage evaluation model includes: inputting the environmental information of the access point device in the first grid and the device information of the access point device into the initial signal coverage evaluation model to obtain first signal coverage prediction information of the access point device in the first grid and second signal coverage prediction information of the access point device in the second grid; determining a first loss value based on the first signal coverage information and the first signal coverage prediction information, the first loss value being used to indicate the difference between the first signal coverage information and the first signal coverage prediction information; determining a second loss value based on the second signal coverage information and the second signal coverage prediction information, the second loss value being used to indicate the difference between the second signal coverage information and the second signal coverage prediction information; and updating the initial signal coverage evaluation model based on the first loss value and the second loss value to obtain the target signal coverage evaluation model.

[0014] In this application, by inputting the environmental information and device information of the access point device within the first grid into the initial signal coverage evaluation model, first signal coverage prediction information of the access point device within the first grid and second signal coverage prediction information of the access point device within the second grid are obtained; based on the first signal coverage information and the first signal coverage prediction information, a first loss value is determined; based on the second signal coverage information and the second signal coverage prediction information, a second loss value is determined; based on the first loss value and the second loss value, the initial signal coverage evaluation model is updated, thereby accurately obtaining the target signal coverage evaluation model.

[0015] Optionally, updating the initial signal coverage assessment model based on the first loss value and the second loss value includes: determining the weights corresponding to the first loss value and the second loss value; determining a target loss value based on the first loss value, the second loss value, the weights of the first loss value and the second loss value, the target loss value being used to indicate the difference between the predicted information and the true information of the initial signal coverage assessment model; and updating the model parameters of the initial signal coverage assessment model based on the target loss value to obtain an updated initial signal coverage assessment model.

[0016] In this application, due to the different methods of acquiring the first and second signal coverage information, the weights corresponding to the first and second loss values ​​are different when updating the model parameters in the initial signal coverage assessment model based on the first and second loss values. Therefore, the weights corresponding to the first and second loss values ​​can be determined, and the target loss value can be identified. Based on the target loss value, the model parameters of the initial signal coverage assessment model can be updated more quickly and reliably, resulting in an updated initial signal coverage assessment model.

[0017] Optionally, determining the weights corresponding to the first loss value and the second loss value includes: obtaining the accuracy of the first signal coverage prediction information and the accuracy of the second signal coverage prediction information; and determining the weights corresponding to the first loss value and the second loss value based on the accuracy of the first signal coverage prediction information and the accuracy of the second signal coverage prediction information.

[0018] In this application, the accuracy rates of the first signal coverage prediction information and the second signal coverage prediction information are obtained; based on the accuracy rates of the first and second signal coverage prediction information, the weights corresponding to the first loss value and the second loss value are determined. In this way, the weights of the first and second loss values ​​can be accurately and reliably determined.

[0019] Optionally, the second grid includes multiple first grids, and the second signal coverage information is used to indicate the average value of the actual signal coverage information of the access point device in the multiple first grids within the second grid. The method further includes: acquiring multiple access probabilities and actual signal coverage information corresponding to each access probability; wherein, the access probability is the access probability of a terminal in the first grid corresponding to the access probability accessing the access point device; and determining the second signal coverage information based on the multiple access probabilities and the actual signal coverage information corresponding to each access probability.

[0020] In this application, during the process of determining the signal coverage information corresponding to the second grid through the measurement report reported by the terminal, if the signal strength of some areas in a region is lower than the receiving sensitivity threshold of the terminal (or user equipment, UE), the terminals in those areas cannot report the measurement report. In this case, the signal coverage information determined solely based on the measurement report cannot comprehensively represent the signal coverage of the entire region. Therefore, it is necessary to combine the terminal access probability within the region to determine the aforementioned second signal coverage information. By acquiring multiple access probabilities and the actual signal coverage information corresponding to each access probability; wherein the access probability is the access probability of a terminal within the first grid corresponding to that access point device; based on multiple access probabilities and the actual signal coverage information corresponding to each access probability, the second signal coverage information can be accurately determined.

[0021] Fourthly, this application provides a model training apparatus, including an acquisition module and a training module. The acquisition module is used to acquire model training data, which includes: environmental information of an access point device within a first grid, device information of the access point device, first signal coverage information, and second signal coverage information. The environmental information is used to indicate the degree to which signal strength is affected within the first grid. The first signal coverage information is the actual signal coverage information of the access point device within the first grid, and the second signal coverage information is the actual signal coverage information of the access point device within the second grid. The first grid is a grid within the second grid. The training module is used to train an initial signal coverage evaluation model based on the model training data to obtain a target signal coverage evaluation model. The target signal coverage evaluation model is used to predict the signal coverage information of the access point device within the first grid based on the environmental information and device information of the access point device within the first grid.

[0022] Optionally, the device further includes a processing module and a determining module; the processing module is used to input the environmental information of the access point device in the first grid and the device information of the access point device into the initial signal coverage evaluation model to obtain first signal coverage prediction information of the access point device in the first grid and second signal coverage prediction information of the access point device in the second grid; the determining module is used to determine a first loss value based on the first signal coverage information and the first signal coverage prediction information, the first loss value being used to indicate the difference between the first signal coverage information and the first signal coverage prediction information; the determining module is also used to determine a second loss value based on the second signal coverage information and the second signal coverage prediction information, the second loss value being used to indicate the difference between the second signal coverage information and the second signal coverage prediction information; the training module is specifically used to update the initial signal coverage evaluation model based on the first loss value and the second loss value to obtain the target signal coverage evaluation model.

[0023] Optionally, the determining module is further configured to determine the weights corresponding to the first loss value and the second loss value; the determining module is further configured to determine a target loss value based on the first loss value, the second loss value, the weights of the first loss value and the weights of the second loss value, the target loss value being used to indicate the difference between the predicted information and the true information of the initial signal coverage evaluation model; the training module is further configured to update the model parameters of the initial signal coverage evaluation model based on the target loss value, thereby obtaining an updated initial signal coverage evaluation model.

[0024] Optionally, the acquisition module is further configured to acquire the accuracy of the first signal coverage prediction information and the accuracy of the second signal coverage prediction information; the determination module is specifically configured to determine the weight corresponding to the first loss value and the weight corresponding to the second loss value based on the accuracy of the first signal coverage prediction information and the accuracy of the second signal coverage prediction information.

[0025] Optionally, the second grid includes multiple first grids, and the second signal coverage information is used to indicate the average value of the actual signal coverage information of the access point device in the multiple first grids within the second grid; the acquisition module is further configured to acquire multiple access probabilities and the actual signal coverage information corresponding to each access probability; wherein, the access probability is the access probability of a terminal in the first grid corresponding to the access probability accessing the access point device; the determination module is further configured to determine the second signal coverage information based on the multiple access probabilities and the actual signal coverage information corresponding to each access probability.

[0026] Fifthly, this application provides an electronic device, including: a processor and a memory configured to store processor-executable instructions; wherein the processor is configured to execute the instructions to implement any of the optional signal evaluation methods in the first aspect above, or to implement any of the optional model training methods in the third aspect above.

[0027] In a sixth aspect, this application provides a computer-readable storage medium storing instructions that, when executed by an electronic device, enable the electronic device to perform any of the optional signal evaluation methods in the first aspect or any of the optional model training methods in the third aspect.

[0028] In a seventh aspect, this application provides a computer program product, which includes computer technology program instructions. When the computer program instructions are executed by a processor, they implement any of the optional signal evaluation methods in the first aspect above, or implement any of the optional model training methods in the third aspect above.

[0029] The signal evaluation method, model training method, apparatus, device, and storage medium provided in this application acquire device information of an access point device and environmental information of the access point device within a first grid. This environmental information indicates the degree of electromagnetic environmental influence on the signal within the first grid. The device information and environmental information of the access point device within the first grid are input into a target signal coverage evaluation model to obtain the signal coverage information of the access point device within the first grid. In this application, by inputting the device information and environmental information of the access point device within the first grid into the target signal coverage evaluation model, the signal coverage situation within the first grid can be obtained accurately and quickly. Attached Figure Description

[0030] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the accompanying drawings used in the description of the embodiments or the prior art will be briefly introduced below.

[0031] Figure 1 This is a schematic diagram of the network architecture of the signal evaluation system provided in the embodiments of this application;

[0032] Figure 2 A schematic flowchart of a signal evaluation method provided in an embodiment of this application;

[0033] Figure 3 A schematic flowchart illustrating a model training method provided in an embodiment of this application;

[0034] Figure 4 A schematic diagram of a cell SSS-RSRP provided for an embodiment of this application;

[0035] Figure 5 Another schematic diagram of cell SSS-RSRP provided in this application embodiment;

[0036] Figure 6 An antenna gain diagram of a first grid provided in an embodiment of this application;

[0037] Figure 7 A 3GPP path loss map including obstacle occlusion information is provided for embodiments of this application;

[0038] Figure 8 An attenuation map of signal attenuation information within a first grid is provided in an embodiment of this application;

[0039] Figure 9 A window diagram corresponding to the position of a first grid provided in an embodiment of this application;

[0040] Figure 10 A functional framework diagram of a Multi-head model provided in this application embodiment;

[0041] Figure 11 A schematic diagram of a model framework provided for an embodiment of this application;

[0042] Figure 12 A schematic diagram of SSS-RSRP rendering provided for an embodiment of this application;

[0043] Figure 13 Another SSS-RSRP-based rendering schematic diagram provided for an embodiment of this application;

[0044] Figure 14 Another SSS-RSRP-based rendering schematic diagram provided for an embodiment of this application;

[0045] Figure 15 A flowchart illustrating another model training method provided in an embodiment of this application;

[0046] Figure 16 A flowchart illustrating another model training method provided in an embodiment of this application;

[0047] Figure 17 This is a schematic diagram of the structure of a signal evaluation device provided in an embodiment of this application;

[0048] Figure 18 This is a schematic diagram of the structure of a model training device provided in an embodiment of this application;

[0049] Figure 19 This is a schematic diagram of another signal evaluation device or model training device provided in the embodiments of this application. Detailed Implementation

[0050] The signal evaluation method, model training method, apparatus, device, and storage medium provided in the embodiments of this application will now be described in detail with reference to the accompanying drawings.

[0051] The terms "first" and "second," etc., in the specification and drawings of this application are used to distinguish different objects, rather than to describe a specific order of objects. For example, "first signal coverage information" and "second signal coverage information," etc., are used to distinguish different signal coverage information, rather than to describe a specific order of signal coverage information.

[0052] Furthermore, the terms “comprising” and “having”, and any variations thereof, used in the description of this application are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or apparatus that includes a series of steps or units is not limited to the steps or units listed, but may optionally include other steps or units not listed, or may optionally include other steps or units inherent to such processes, methods, products, or apparatus.

[0053] It should be noted that in the embodiments of this application, the words "exemplary" or "for example" are used to indicate examples, illustrations, or explanations. Any embodiment or design scheme described as "exemplary" or "for example" in the embodiments of this application should not be construed as being more preferred or advantageous than other embodiments or design schemes. Specifically, the use of the words "exemplary" or "for example" is intended to present the relevant concepts in a specific manner.

[0054] The term "and / or" as used in this application includes using either one of two methods or using both methods simultaneously.

[0055] In the description of this application, unless otherwise stated, "a plurality of" means two or more.

[0056] Wireless network services are a crucial component of modern communication services, and assessing wireless network signal coverage is the most fundamental and critical step in network planning. The accuracy of assessing wireless network signal coverage within a specific area directly determines wireless network performance indicators, user experience, and resource utilization efficiency. On the other hand, due to the rapidly increasing demand for wireless network capacity, cellular networks are increasingly using mid-to-high frequency bands (2-6 GHz) for communication equipment. Compared to low-frequency signals, these bands are more susceptible to environmental influences during propagation, posing new challenges to wireless signal coverage assessment.

[0057] Currently, user equipment (UEs) accessing a cellular network periodically send a measurement report (MR) to their access point. This report includes the secondary synchronization signal reference signal received power (SSS-RSRP) value, measured based on the secondary synchronization signal. However, due to user movement, insufficient positioning accuracy, and data access restrictions, only an average SSS-RSRP value at a 50-meter grid level over a specific period is typically obtained. This grid size exceeds the floor area of ​​most buildings, making this granular SSS-RSRP data unsuitable for accurately assessing wireless signal coverage (e.g., determining whether coverage blind spots exist inside buildings or between buildings). Furthermore, UEs can only generate a measurement report for a specific access point after connecting to it. Therefore, the SSS-RSRP value in the measurement report collected from a specific access point is also affected by the UE's access probability, failing to objectively reflect the wireless signal coverage around that access point.

[0058] In addition, by having surveyors carry measuring equipment to measure the signal strength at multiple locations within an area, the signal coverage of that area can be determined.

[0059] However, the above methods, which rely on manual measurement to determine the signal coverage of an area, have the problems of excessive consumption of manpower and resources, and may not be able to obtain the signal coverage information of the area in a timely manner; or there may be blind spots or low accuracy in determining the signal coverage of an area based on existing reported measurement reports.

[0060] Based on this, embodiments of this application provide a signal evaluation method, a model training method, an apparatus, a device, and a storage medium. By inputting the device information of the access point device and the environmental information of the access point device within the first grid into the target signal coverage evaluation model, the signal coverage within the first grid can be obtained accurately and quickly.

[0061] The signal evaluation method, model training method, apparatus, device, and storage medium provided in this application can be applied to signal evaluation systems, such as... Figure 1 As shown, the signal evaluation system includes an access point device 101 and a terminal 102. Typically, in practical applications, the connection between these devices can be wireless. To conveniently and intuitively illustrate the connection relationships between the devices, Figure 1 Solid lines are used to represent the meaning.

[0062] The access point device 101 is used to transmit wireless signals so that the terminal 102 can access the wireless network based on the access point device 101.

[0063] Terminal 102 is used to access the wireless network via access point device 101; or to send a measurement report to access point device 101.

[0064] Optionally, the measurement report may include the reference signal received power detected by terminal 102 based on the subsynchronous signal measurement.

[0065] For example, the access point device 101 mentioned above may include base stations, evolved node base stations (eNBs), next-generation node base stations (gNBs), new radio eNBs, macro base stations, micro base stations, high-frequency base stations or transmission and reception points (TRPs), non-3GPP access networks (such as WiFi), and / or non-3GPP interworking functions (N3IWFs), etc.

[0066] For example, the electronic device executing the model training method provided in this application embodiment, or the aforementioned terminal 102, can be a mobile phone, tablet computer, desktop computer, laptop computer, handheld computer, notebook computer, ultra-mobile personal computer (UMPC), netbook, as well as cellular phone, personal digital assistant (PDA), augmented reality (AR) / virtual reality (VR) device. This application embodiment does not impose special limitations on the specific form of the electronic device. It can interact with the user through one or more methods such as a keyboard, touchpad, touchscreen, remote control, voice interaction, or handwriting device.

[0067] like Figure 2 As shown, the signal evaluation method provided in this application embodiment may include S201-S202.

[0068] S201. Obtain the device information of the access point device and the environmental information of the access point device within the first grid.

[0069] Among them, environmental information is used to indicate the degree to which the signal is affected by the electromagnetic environment within the first grid.

[0070] In one possible implementation, the device information of the access point device is stored in the access point device; the environmental information of the access point device within the first grid can be measured by an environmental measurement tool.

[0071] S202. Input the device information of the access point device and the environmental information of the access point device in the first grid into the target signal coverage evaluation model to obtain the signal coverage information of the access point device in the first grid.

[0072] In one possible implementation, the target signal coverage assessment model is obtained by training an initial signal coverage assessment model based on model training parameters. The model training parameters include: signal assessment method, model training method, environmental information of the access point device within the first grid, device information of the access point device, first signal coverage information, and second signal coverage information. The environmental information is used to indicate the degree to which the signal strength is affected within the first grid. The first signal coverage information is the actual signal coverage information of the access point device within the first grid, and the second signal coverage information is the actual signal coverage information of the access point device within the second grid. The first grid is a grid within the second grid.

[0073] In this embodiment of the application, by inputting the device information of the access point device and the environmental information of the access point device within the first grid into the target signal coverage evaluation model, the signal coverage within the first grid can be obtained accurately and quickly.

[0074] Furthermore, embodiments of this application also provide a model training method for obtaining the aforementioned target signal coverage evaluation model. For example... Figure 3 As shown, the model training methods include S301-S302:

[0075] S301. Obtain model training data.

[0076] The model training data includes: environmental information of the access point device within the first grid, device information of the access point device, first signal coverage information, and second signal coverage information. The environmental information is used to indicate the degree to which the signal strength is affected within the first grid. The first signal coverage information is the actual signal coverage information of the access point device within the first grid, and the second signal coverage information is the actual signal coverage information of the access point device within the second grid. The first grid is a grid within the second grid.

[0077] In some embodiments, the device information of the access point device includes the location of the signal transmitting device (such as an antenna), the frequency of the transmitted signal, and the antenna pattern of the signal transmitting device.

[0078] Alternatively, the antenna gain map of the first grid (or signal receiving point) can be determined based on the device information of the access point device to indicate the signal gain information of the antenna within the first grid.

[0079] In some embodiments, the environmental information of the access point device within the first grid may be geographic environmental information, including an obstacle map (such as a building map) within the first grid and the altitude of the first grid.

[0080] It should be understood that the obstacle map includes information such as the shape, height, and material of the obstacles; based on the obstacle map, the extent of signal obstruction within the first grid can be determined. The elevation of the first grid is used to determine the signal attenuation within the first grid. Thus, based on the signal obstruction within the first grid and the signal attenuation within the first grid, the degree of signal impact within the first grid can be determined.

[0081] Understandably, based on the characteristics of wireless communication equipment and the principle of electromagnetic wave propagation, the antenna gain at the first grid (or the receiving device or terminal), the building obstruction effect within the first grid, and the signal attenuation effect within the first grid have a significant impact on the signal strength (or SSS-RSRP) at the receiving point.

[0082] In one possible implementation, the first signal coverage information is obtained through field testing.

[0083] In one possible implementation, the first signal coverage information is obtained through simulation calculation.

[0084] In one possible implementation, the second signal coverage information is obtained from the reference signal received power in a measurement report sent by a terminal within the second grid.

[0085] For example, a terminal (or user equipment) in the second grid reports a measurement report including the reference signal received power to the access point equipment (or base station), and the reference signal received power can be extracted from the measurement report stored on the network management side of the base station.

[0086] In some embodiments, the first signal coverage information is referred to as fine-grained signal coverage information, and the second signal coverage information is referred to as coarse-grained signal coverage information.

[0087] In some embodiments, the second signal coverage information is referred to as the average value of multiple coarse-grained signal coverage information within the second grid.

[0088] In one possible implementation, the second grid includes multiple first grids, and the second signal coverage information is used to indicate the average value of the actual signal coverage information of the access point device within the multiple first grids in the second grid. In this case, the second signal coverage information can be determined by: acquiring multiple access probabilities and the actual signal coverage information corresponding to each access probability; wherein the access probability is the probability of a terminal accessing the access point device within the first grid corresponding to the access probability; and determining the second signal coverage information based on the multiple access probabilities and the actual signal coverage information corresponding to each access probability.

[0089] It should be understood that, in the process of determining the signal coverage information corresponding to the second grid through the measurement report reported by the terminal, if the signal strength of some areas in a region is lower than the receiving sensitivity threshold of the terminal (or user equipment, UE), the terminal in that area cannot report the measurement report. In this case, the signal coverage information determined solely based on the measurement report cannot fully represent the signal coverage of the entire region. It is necessary to combine the terminal access probability in the region to determine the aforementioned second signal coverage information.

[0090] It is understandable that, for a terminal in a region, when there are signals from multiple access point devices in the region, the terminal will access the access point device with the highest signal (or access the cellular network).

[0091] For example, assuming the signal coverage information is SSS-RSRP, using RSRP (x,y) This represents the actual SSS-RSRP value at position (x,y), TH represents the receiver sensitivity threshold of the terminal, and M... rsrp Let SS represent the average value of SSS-RSRP in the MR collected at a specific receiving point. Then, the access probability of the terminal at the (x,y) location satisfies the following formula (1):

[0092]

[0093] Assume there are M second grid cells in region S, and the m-th second grid cell is represented by S. m This indicates that the average RSRP value within the m-th grid in the measurement report is represented by RSRP. m In theory, RSRP m It can be calculated based on the following two-dimensional integral, satisfying the following formula (2):

[0094]

[0095] Using the finite element method, assume S m There are K terminals. The position of the kth terminal is represented by (x). k ,y k ) indicates that RSRP mapproximation The following formula (3) must be satisfied:

[0096]

[0097] Use Δ m Represents the theoretical value RSRP m and approximate value For ease of calculation, the difference between them can be treated as a random variable, which follows a Gaussian distribution with mean μ and variance σ. The Gaussian distribution satisfies formula (4):

[0098]

[0099] In this way, based on the SSS-RSRP in the test report and the access probability of the terminal, the estimated value of the second signal coverage information can be obtained. Based on the difference between the theoretical value and the estimated value, the theoretical value of the second signal coverage information can be determined, thus obtaining a more accurate second signal coverage information.

[0100] Furthermore, the second signal coverage information can also be determined based on the first signal coverage information, meaning there is a correlation between the first signal coverage information and the second signal coverage information. Thus, the second signal coverage information determined based on the measurement report can also be used to reverse-engineer the first signal coverage information.

[0101] In this embodiment of the application, the output of the initial signal coverage evaluation model includes first signal coverage prediction information and second signal coverage prediction information, wherein the second signal coverage prediction information can be determined based on the first signal coverage prediction information.

[0102] In one possible implementation, the second signal prediction information can be determined based on the first signal prediction information by using a custom neural network layer.

[0103] For example, the theoretical values ​​of SSS-RSRP in the prediction information {RSRP1, RSRP2…RSRP} are covered by the first signal. K}, receiver sensitivity threshold TH, mean SSS-RSRP in the measurement report M rsrp As input to a custom neural network layer, the second signal covers the average value of SSS-RSRP in the prediction information and the difference Δ between the theoretical and estimated values. m The mean value b is used as the output of the custom neural network layer.

[0104] Specifically, a trainable parameter b is defined in the custom neural network layer, the theoretical value of SSS-RSRP is calculated based on the above formula (2), the estimated value of the second signal coverage prediction information is calculated based on the above formula (3), and the second signal coverage prediction information is determined based on the estimated value of the second signal coverage prediction information and b.

[0105] It should be understood that the above formula demonstrates the functional relationship between the SSS-RSRP at the first grid level and the SSS-RSRP at the second grid level. The determination of the SSS-RSRP at the second grid level using access probability and measurement reports is based on statistical characteristic analysis.

[0106] Optionally, the position point (x) in the above formula can be... k ,y k The first grid cell can be considered as a 5m grid cell. For example, if the first grid cell is 5m, then the position point (x) in the above formula can be considered as the first grid cell. k ,y k Consider it as the first grid S k 5 .

[0107] Optionally, the second grid is a 50m grid and the first grid is a 5m grid.

[0108] In addition, such as Figure 4 As shown in the illustration, this application also provides a schematic diagram of a cell SSS-RSRP to illustrate the aforementioned SSS-RSRP, including a single-cell SSS-RSRP histogram. For example... Figure 5 The diagram shown illustrates another cell SSS-RSRP schematic provided in this application embodiment, including a multi-cell SSS-RSRP histogram. Both the single-cell and multi-cell SSS-RSRP histograms are determined based on the SSS-RSRP from measurement reports from terminals within the cell. The single-cell SSS-RSRP histogram exemplarily shows the histogram corresponding to the SSS-RSRP collected from a single cell (one wireless cellular access point), with SSS-RSRP values ​​ranging from [-125, -60]. The SSS-RSRP values ​​for both single-cell and multi-cell cells show a significant frequency drop when below -100dBm, indicating a lower number of terminals accessing the wireless access point at values ​​below -100dBm. Therefore, there is a problem of inaccurate signal coverage information determined based on measurement reports.

[0109] S302. Train the initial signal coverage evaluation model based on the model training data to obtain the target signal coverage evaluation model.

[0110] The target signal coverage assessment model is used to predict the signal coverage information of the access point device within the first grid based on the environmental information and device information of the access point device within the first grid.

[0111] In one possible implementation, the target signal coverage evaluation model is not obtained through machine learning from an AI model.

[0112] In one possible implementation, training the initial signal coverage assessment model based on model training data may include inputting a grayscale image of the environment information of the access point device within the first grid, a grayscale image of the device information of the access point device, and a grayscale image of the position of the first grid into the initial signal coverage assessment model.

[0113] For example, the grayscale image corresponding to the environmental information of the access point device within the first grid is the antenna gain image of the first grid; the grayscale image corresponding to the device information of the access point device is a 3GPP path loss image containing obstacle occlusion information and an attenuation image containing signal attenuation information within the first grid; the grayscale image corresponding to the position of the first grid is the window image corresponding to the position of the first grid.

[0114] Optionally, the grayscale value of the window image corresponding to the first grid position is 1 at the first grid position, and the grayscale value of the exotic region is 0.

[0115] Optionally, the grayscale image is a grayscale image with dimensions of 129 pixels × 129 pixels.

[0116] Optionally, each pixel in the grayscale image above can be used to represent a 5m grid. The first grid is one of these.

[0117] Optionally, the grayscale image corresponding to the environmental information of the access point device within the first grid can be obtained by standardizing the environmental information.

[0118] Optionally, the first grid is an indoor area, and the attenuation map of signal attenuation information in the first grid can be an Outdoor2indoor attenuation map that includes indoor signal attenuation information.

[0119] For example, such as Figure 6 The image shown is an antenna gain diagram of a first grid provided in an embodiment of this application; as shown... Figure 7 As shown, this is a 3GPP path loss map including obstacle occlusion information provided in an embodiment of this application; Figure 8 As shown, this is an attenuation map of signal attenuation information within a first grid according to an embodiment of this application; Figure 9 The image shown is a window diagram corresponding to the position of a first grid provided in an embodiment of this application.

[0120] In one possible implementation, the initial signal coverage evaluation model is a multi-head model built on a convolutional neural network (CNN).

[0121] Optionally, the multi-head model also includes custom neural network layers.

[0122] In one possible implementation, the second grid includes multiple first grids. The antenna gain of a first grid, obstacle obstruction information of a first grid, signal attenuation information of a first grid, and the position of a first grid are determined as a set of electromagnetic environment information for the first grid. The antenna gain map of a first grid, the 3GPP path loss map of a first grid containing obstacle obstruction information, the attenuation map of a first grid containing signal attenuation information, and the window map corresponding to the position of a first grid are determined as a grayscale map corresponding to a set of electromagnetic environment information.

[0123] In one possible implementation, in the Multi-head model, each set of electromagnetic environment information corresponds to a grayscale image with two convolutional layers and one fully connected layer containing multiple neurons. A aggregation layer generates fine-grained signal coverage information based on the output of each fully connected layer, and a custom neural network layer generates coarse-grained signal coverage information. Then, based on the first signal coverage information and the fine-grained signal coverage information, a loss value corresponding to the fine-grained signal coverage information is obtained; based on the second signal coverage information and the coarse-grained signal coverage information, a loss value corresponding to the coarse-grained signal coverage information is obtained; and based on the loss values ​​corresponding to the coarse-grained and fine-grained signal coverage information, the model parameters of the Multi-head model are updated to obtain the updated Multi-head model. If the loss values ​​corresponding to the fine-grained and coarse-grained signal coverage information generated by the updated Multi-head model are less than or equal to a loss value threshold, the updated Multi-head model is determined as the target signal coverage evaluation model.

[0124] In one possible implementation, the training process of the multi-head model employs iterative optimization using adaptive moment estimation (Adam).

[0125] Optionally, the learning rate during the Adam iterative optimization process is set to 2.5e-5.

[0126] In some embodiments, a set of grayscale images is called a dataset, and the four grayscale images in a dataset are called the four inputs of the Multi-head model, denoted as Input1, Input2, Input3, and Input4, respectively. Fine-grained signal coverage information and coarse-grained signal coverage information are denoted as Output1 and Output2, respectively. Output1 is the output of the aggregation layer (which can be called concatenate) and also the input of the custom neural network layer (which can be called customdenselayer), with a data format of (None, 4). Output2 is the output of the custom neural network layer (which can be called customdenselayer), with a data format of (None, 1).

[0127] For example, such as Figure 10 The diagram shown is a functional framework diagram of a Multi-head model provided in an embodiment of this application, including multiple sets of electromagnetic environment information, grayscale images corresponding to each set of electromagnetic environment information, a first convolutional layer corresponding to each set of grayscale images, a max pooling result corresponding to each set of grayscale images, a second convolutional layer corresponding to each set of grayscale images, a fully connected layer corresponding to each set of grayscale images, an aggregation layer, and a custom neural network layer.

[0128] For example, assuming there are four sets of electromagnetic environment information, a multi-head CNN model is built based on the TensorFlow machine learning library. The model framework is as follows: Figure 11 As shown, this includes the input, output, name, type, input data format, and output data format of each neural network layer. Taking the neural network layer named "Input_5" as an example, the meaning of the parameters for each layer is explained. The input data format for "Input_5" is (batch size, 128, 128, 4), where "batch size" represents the input batch size (optionally, "none" indicates dynamically determined batch size), "128, 128" represents the input data size as 128 pixels × 128 pixels, and "4" represents the number of input channels as 4. The meaning of the parameters for other neural network layers can be derived similarly.

[0129] In addition, such as Figure 12 As shown, this application embodiment also provides a rendering diagram based on SSS-RSRP, used to visualize coarse-grained SSS-RSRP obtained from measurement reports; as shown Figure 13The diagram shown is another rendering illustration based on SSS-RSRP provided in this application embodiment, used to visualize the fine-grained SSS-RSRP generated by the Multi-head CNN model; as shown Figure 14 The diagram shown is another SSS-RSRP rendering schematic provided in this application embodiment, including a superimposed rendering of a fine-grained SSS-RSRP rendering and a coarse-grained SSS-RSRP rendering. The superimposed rendering of the fine-grained and coarse-grained SSS-RSRP renderings is obtained by superimposing the latitude and longitude of the two renderings respectively. It can be seen that the fine-grained SSS-RSRP data calculated by the model conforms to the distribution of the measurement report data, and the missing areas in the measurement report can be determined based on the fine-grained SSS-RSRP data, i.e., there are coverage blind spots in the coarse-grained SSS-RSRP rendering. Figure 14 (The area within the Chinese frame).

[0130] In this embodiment, since the first grid is a grid within the second grid, the first signal coverage information can more accurately reflect the signal coverage within the first grid than the second signal coverage information, and there is a correlation between the first and second signal coverage information. Because the target signal coverage evaluation model is used to predict the signal coverage information of the access point device within the first grid based on the environmental information and device information of the access point device within the first grid—that is, the first and second signal coverage information serve as the validation set for the initial signal coverage evaluation model during training—the initial signal coverage evaluation model can be trained using both the more accurate first signal coverage information and the less accurate second signal coverage information, resulting in a higher prediction accuracy for the target signal coverage evaluation model.

[0131] Furthermore, since the grid corresponding to the second signal coverage information is larger than that corresponding to the first signal coverage information, obtaining the second signal coverage information is more convenient and faster; conversely, obtaining the first signal coverage information is more difficult. When the amount of available first signal coverage information is limited, the prediction accuracy of the trained model is low. Therefore, this embodiment of the application improves the prediction accuracy of the trained target signal coverage evaluation model by adding second signal coverage information that is related to the first grid during the training process.

[0132] In one possible implementation, combining Figure 3 The illustrated embodiments, such as Figure 15As shown, in step S302 above, the initial signal coverage evaluation model is trained based on the model training data to obtain the target signal coverage evaluation model, including steps S1501-S1503:

[0133] S1501. Input the environmental information and equipment information of the access point device in the first grid into the initial signal coverage evaluation model to obtain the first signal coverage prediction information of the access point device in the first grid and the second signal coverage prediction information of the access point device in the second grid.

[0134] S1502. Based on the first signal coverage information and the first signal coverage prediction information, determine the first loss value; based on the second signal coverage information and the second signal coverage prediction information, determine the second loss value.

[0135] The first loss value is used to indicate the difference between the first signal coverage information and the first signal coverage prediction information; the second loss value is used to indicate the difference between the second signal coverage information and the second signal coverage prediction information.

[0136] In one possible implementation, the determination of the first and second loss values ​​can be achieved using the mean squared error (MSE) loss function.

[0137] S1503. Based on the first loss value and the second loss value, update the initial signal coverage evaluation model to obtain the target signal coverage evaluation model.

[0138] In one possible implementation, the initial signal coverage evaluation model is updated based on the first loss value and the second loss value to obtain the target signal coverage evaluation model. It may also include: evaluating the updated initial signal coverage evaluation model based on the mean square error and absolute error loss functions and the test set data to obtain the evaluation score corresponding to the updated initial signal coverage evaluation model. The evaluation score is used to indicate the accuracy of the signal coverage prediction information generated by the updated initial signal coverage evaluation model.

[0139] Optionally, if the evaluation score is lower than the score threshold, it indicates that the accuracy of the signal coverage prediction information generated by the updated initial signal coverage evaluation model is low. In this case, the initial signal coverage evaluation model is updated again.

[0140] Optionally, multiple updated evaluation scores can be used. If an evaluation score below a score threshold is greater than a numerical threshold, it indicates that the updated initial signal coverage evaluation model may be overfitting. In this case, the training process is terminated and the model training data is updated, increasing the size and types of model training data (for example, electromagnetic environment information of the first or second grid in different regions can be added).

[0141] Optionally, if the evaluation score is higher than the score threshold, it indicates that the signal coverage prediction information generated by the updated initial signal coverage evaluation model is highly accurate. In this case, training is terminated and the updated initial signal coverage evaluation model is determined as the target signal coverage evaluation model.

[0142] It should be understood that by inputting the environmental information and device information of the access point device within the first grid into the initial signal coverage assessment model, the first signal coverage prediction information of the access point device within the first grid and the second signal coverage prediction information of the access point device within the second grid are obtained; based on the first signal coverage information and the first signal coverage prediction information, a first loss value is determined; based on the second signal coverage information and the second signal coverage prediction information, a second loss value is determined; based on the first loss value and the second loss value, the initial signal coverage assessment model is updated, thereby accurately obtaining the target signal coverage assessment model.

[0143] In one possible implementation, combining Figure 15 The illustrated embodiments, such as Figure 16 As shown, based on the first loss value and the second loss value, the initial signal coverage evaluation model is updated to obtain the target signal coverage evaluation model, including S1601-S1603:

[0144] S1601. Determine the weights corresponding to the first loss value and the weights corresponding to the second loss value.

[0145] It should be understood that, due to the different acquisition methods of the first signal coverage information and the second signal coverage information, the weights corresponding to the first loss value and the second loss value are different when updating the model parameters in the initial signal coverage evaluation model based on the first loss value and the second loss value. At this time, the weights corresponding to the first loss value and the second loss value can be determined.

[0146] In one possible implementation, the weights corresponding to the first loss value and the second loss value are determined by: obtaining the accuracy rates of the first signal coverage prediction information and the second signal coverage prediction information; and determining the weights corresponding to the first loss value and the second loss value based on the accuracy rates of the first and second signal coverage prediction information. In this way, the weights of the first loss value and the second loss value can be determined accurately and reliably.

[0147] In one possible implementation, the weights corresponding to the first loss value and the second loss value can also be determined based on the accuracy of the first signal coverage information and the accuracy of the second signal coverage information.

[0148] In one possible implementation, the accuracy of the first signal coverage information can also be determined based on the method of acquiring the first signal coverage information.

[0149] For example, assuming that the first signal coverage information is obtained by manual on-site collection and the second signal coverage information is obtained by measurement reports sent by the terminal, the weights corresponding to the first loss value and the second loss value can be determined based on the reliability of the first signal coverage information when it is manually collected on-site and the reliability of the second signal coverage information when it is sent by the terminal.

[0150] For example, the weight corresponding to the first loss value is w1, the weight corresponding to the second loss value is w2, and w1+w2=1.

[0151] For example, assuming the first signal coverage information is calculated using a ray tracing model, we can determine that w1 = w2 = 0.5. Alternatively, assuming the first signal coverage information is obtained through manual field measurement, we can determine that w1 = 0.16 and w2 = 0.2.

[0152] For example, the initial signal coverage evaluation model can be trained using the training parameters shown in Table 1:

[0153] Table 1

[0154]

[0155] S1602. Determine the target loss value based on the first loss value, the second loss value, the weight of the first loss value, and the weight of the second loss value.

[0156] The target loss value is used to indicate the difference between the predicted information and the actual information of the initial signal coverage evaluation model.

[0157] In one possible implementation, the target loss value is determined by the sum of the product of the first loss value and its weight, and the product of the second loss value and its weight.

[0158] S1603. Based on the target loss value, update the model parameters of the initial signal coverage assessment model to obtain the updated initial signal coverage assessment model.

[0159] In one possible implementation, the model parameters of the initial signal coverage evaluation model can be updated based on gradient descent, according to the gradient of the decrease in the target loss value.

[0160] In this embodiment, due to the different acquisition methods of the first and second signal coverage information, the weights corresponding to the first and second loss values ​​are different when updating the model parameters in the initial signal coverage evaluation model based on the first and second loss values. Therefore, the weights corresponding to the first and second loss values ​​can be determined, and the target loss value can be identified. Based on the target loss value, the model parameters of the initial signal coverage evaluation model can be updated more quickly and reliably, resulting in an updated initial signal coverage evaluation model.

[0161] For example, this application also provides a model training process, including: analyzing the statistical features of SSS-RSRP in measurement report data; deriving the conversion formula of coarse and fine granular SSS-RSRP; designing a custom neural network layer according to the conversion formula; building a Multi-head CNN neural network framework; generating input and output datasets; training the model and testing the coverage evaluation results.

[0162] This application embodiment can divide the model training device, signal evaluation device, etc., into functional modules according to the above method examples. For example, each function can be divided into its own functional module, or two or more functions can be integrated into one processing module. The integrated module can be implemented in hardware or as a software functional module. It should be noted that the module division in this application embodiment is illustrative and only represents one logical functional division; other division methods may be used in actual implementation.

[0163] When dividing each function into modules according to its corresponding function. Figure 17 A possible structural schematic diagram of the signal evaluation device involved in the above embodiments is shown, such as... Figure 17 As shown, the signal evaluation device may include an acquisition module 1701 and a processing module 1702.

[0164] The acquisition module 1701 is used to acquire device information of the access point device and environmental information of the access point device within a first grid, wherein the environmental information is used to indicate the degree to which the signal strength is affected within the first grid.

[0165] The processing module 1702 is used to input the device information of the access point device and the environmental information of the access point device in the first grid into the target signal coverage evaluation model to obtain the signal coverage information of the access point device in the first grid.

[0166] Optionally, the target signal coverage assessment model is obtained by training an initial signal coverage assessment model based on model training parameters. The model training parameters include: environmental information of the access point device in the first grid, device information of the access point device, first signal coverage information, and second signal coverage information. The environmental information is used to indicate the degree of signal impact in the first grid. The first signal coverage information is the actual signal coverage information of the access point device in the first grid. The second signal coverage information is the actual signal coverage information of the access point device in the second grid. The first grid is smaller than the second grid.

[0167] When dividing each function into modules according to its corresponding function. Figure 18 A possible structural schematic diagram of the model training device involved in the above embodiments is shown, such as... Figure 18 As shown, the model training device may include an acquisition module 1801 and a training module 1802.

[0168] The acquisition module 1801 is used to acquire model training data. The model training data includes: environmental information of the access point device in the first grid, device information of the access point device, first signal coverage information, and second signal coverage information. The environmental information is used to indicate the degree of influence of signal strength in the first grid. The first signal coverage information is the actual signal coverage information of the access point device in the first grid. The second signal coverage information is the actual signal coverage information of the access point device in the second grid. The first grid is a grid within the second grid.

[0169] Training module 1802 is used to train the initial signal coverage evaluation model based on model training data to obtain the target signal coverage evaluation model. The target signal coverage evaluation model is used to predict the signal coverage information of the access point device in the first grid based on the environmental information and device information of the access point device within the first grid.

[0170] Optionally, the device further includes a processing module 1803 and a determination module 1804.

[0171] The processing module 1803 is used to input the environmental information and equipment information of the access point device in the first grid into the initial signal coverage evaluation model to obtain the first signal coverage prediction information of the access point device in the first grid and the second signal coverage prediction information of the access point device in the second grid.

[0172] The determination module 1804 is used to determine a first loss value based on the first signal coverage information and the first signal coverage prediction information. The first loss value is used to indicate the difference between the first signal coverage information and the first signal coverage prediction information.

[0173] The determining module 1804 is further configured to determine a second loss value based on the second signal coverage information and the second signal coverage prediction information, wherein the second loss value is used to indicate the difference between the second signal coverage information and the second signal coverage prediction information.

[0174] Training module 1802 is specifically used to update the initial signal coverage evaluation model based on the first loss value and the second loss value to obtain the target signal coverage evaluation model.

[0175] Optionally, the determining module 1804 is also used to determine the weight corresponding to the first loss value and the weight corresponding to the second loss value.

[0176] The determination module 1804 is further configured to determine a target loss value based on the first loss value, the second loss value, the weight of the first loss value, and the weight of the second loss value. The target loss value is used to indicate the difference between the predicted information and the true information of the initial signal coverage evaluation model.

[0177] The training module 1802 is also specifically used to update the model parameters of the initial signal coverage evaluation model based on the target loss value, so as to obtain the updated initial signal coverage evaluation model.

[0178] Optionally, the acquisition module 1801 is also used to acquire the accuracy of the first signal coverage prediction information and the accuracy of the second signal coverage prediction information.

[0179] The determination module 1804 is specifically used to determine the weights corresponding to the first loss value and the second loss value based on the accuracy of the first signal coverage prediction information and the accuracy of the second signal coverage prediction information.

[0180] Optionally, the second grid includes a plurality of first grids, and the second signal coverage information is used to indicate the average value of the actual signal coverage information of the access point device within the plurality of first grids in the second grid.

[0181] The acquisition module 1801 is also used to acquire multiple access probabilities and the actual signal coverage information corresponding to each access probability. The access probability is the probability that a terminal within the first grid corresponding to the access probability will access the access point device.

[0182] The determining module 1804 is also used to determine the second signal coverage information based on multiple access probabilities and the actual signal coverage information corresponding to each access probability.

[0183] When using integrated units, Figure 19 A possible structural schematic diagram of the signal evaluation device or model training device involved in the above embodiments is shown. For example... Figure 19As shown, the signal evaluation device or model training device may include a processing module 1901 and a communication module 1902. The processing module 1901 can be used to control and manage the operation of the signal evaluation device or model training device. The communication module 1902 can be used to support communication between the signal evaluation device or model training device and other entities. Optionally, as... Figure 19 As shown, the signal evaluation device or model training device may further include a storage module 1903 for storing the program code and data of the signal evaluation device or model training device.

[0184] The processing module 1901 can be a processor or a controller. The communication module 1902 can be a transceiver, transceiver circuit, or communication interface, etc. The storage module 1903 can be a memory.

[0185] In this configuration, when the processing module 1901 is a processor, the communication module 1902 is a transceiver, and the storage module 1903 is a memory, the processor, transceiver, and memory can be connected via a bus. The bus can be a PCI bus or an EISA bus, etc. The bus can be categorized as an address bus, data bus, control bus, etc.

[0186] It should be understood that in the various embodiments of this application, the order of the above-mentioned processes does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0187] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0188] Those skilled in the art will understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0189] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0190] In the above embodiments, implementation can be achieved, in whole or in part, through software, hardware, firmware, or any combination thereof. When implemented using software programs, implementation can be, in whole or in part, in the form of a computer program product. This computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the processes or functions described in the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, fiber optic, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium accessible to a computer or a data storage device containing one or more servers, data centers, etc., that can be integrated with the medium. The available media can be magnetic media (e.g., floppy disks, hard disks, magnetic tapes), optical media (e.g., DVDs), or semiconductor media (e.g., solid-state disks (SSDs)).

[0191] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A signal evaluation method, characterized in that, The method includes: Obtain device information of the access point device and environmental information of the access point device within a first grid, wherein the environmental information is used to indicate the degree to which signal strength is affected within the first grid; The device information of the access point device and the environmental information of the access point device in the first grid are input into the target signal coverage evaluation model to obtain the signal coverage information of the access point device in the first grid. The target signal coverage evaluation model is obtained by training an initial signal coverage evaluation model based on model training parameters. The model training parameters include: a signal evaluation method, environmental information of the access point device within a first grid, device information of the access point device, first signal coverage information, and second signal coverage information. The environmental information indicates the degree to which signal strength is affected within the first grid. The first signal coverage information is the actual signal coverage information of the access point device within the first grid, where the first grid is a grid within a second grid. The second signal coverage information is the actual signal coverage information of the access point device within the second grid. The second signal coverage information indicates the average value of the actual signal coverage information of the access point device in multiple first grids within the second grid. The second signal coverage information is determined based on multiple access probabilities and the actual signal coverage information corresponding to each access probability. The access probability is the probability that a terminal within the first grid corresponding to the access probability will access the access point device.

2. A model training method, characterized in that, The method includes: The model training data includes: environmental information of the access point device within a first grid, device information of the access point device, first signal coverage information, and second signal coverage information. The environmental information indicates the degree to which signal strength is affected within the first grid. The first signal coverage information is the actual signal coverage information of the access point device within the first grid, and the second signal coverage information is the actual signal coverage information of the access point device within a second grid. The first grid is a grid within the second grid. The second signal coverage information indicates the average value of the actual signal coverage information of the access point device in multiple first grids within the second grid. The second signal coverage information is determined based on multiple access probabilities and the actual signal coverage information corresponding to each access probability. The access probability is the probability that a terminal within the first grid corresponding to the access probability will access the access point device. The initial signal coverage evaluation model is trained based on the model training data to obtain the target signal coverage evaluation model; the target signal coverage evaluation model is used to predict the signal coverage information of the access point device in the first grid based on the environmental information of the access point device in the first grid and the device information of the access point device.

3. The method according to claim 2, characterized in that, The step of training the initial signal coverage evaluation model based on the model training data to obtain the target signal coverage evaluation model includes: The environmental information and device information of the access point device in the first grid are input into the initial signal coverage evaluation model to obtain the first signal coverage prediction information of the access point device in the first grid and the second signal coverage prediction information of the access point device in the second grid. Based on the first signal coverage information and the first signal coverage prediction information, a first loss value is determined, which indicates the difference between the first signal coverage information and the first signal coverage prediction information; based on the second signal coverage information and the second signal coverage prediction information, a second loss value is determined, which indicates the difference between the second signal coverage information and the second signal coverage prediction information. Based on the first loss value and the second loss value, the initial signal coverage evaluation model is updated to obtain the target signal coverage evaluation model.

4. The method according to claim 3, characterized in that, Based on the first loss value and the second loss value, the initial signal coverage evaluation model is updated, including: Determine the weight corresponding to the first loss value and the weight corresponding to the second loss value; Based on the first loss value, the second loss value, the weight of the first loss value, and the weight of the second loss value, a target loss value is determined, which is used to indicate the difference between the predicted information and the true information of the initial signal coverage evaluation model. Based on the target loss value, the model parameters of the initial signal coverage assessment model are updated to obtain the updated initial signal coverage assessment model.

5. The method according to claim 4, characterized in that, Determining the weights corresponding to the first loss value and the second loss value includes: Obtain the accuracy of the first signal coverage prediction information and the accuracy of the second signal coverage prediction information; Based on the accuracy of the first signal coverage prediction information and the accuracy of the second signal coverage prediction information, the weights corresponding to the first loss value and the second loss value are determined.

6. The method according to claim 2, characterized in that, The second grid includes a plurality of the first grids, and the second signal coverage information is used to indicate the average value of the actual signal coverage information of the access point device within the plurality of the first grids in the second grid. The method further includes: Acquire multiple access probabilities and the actual signal coverage information corresponding to each access probability; wherein, the access probability is the access probability of a terminal in the first grid corresponding to the access probability accessing the access point device; The second signal coverage information is determined based on multiple access probabilities and the actual signal coverage information corresponding to each access probability.

7. A model training device, characterized in that, Includes an acquisition module, a determination module, and a training module: The acquisition module is used to acquire model training data, which includes: environmental information of the access point device within a first grid, device information of the access point device, first signal coverage information, and second signal coverage information. The environmental information is used to indicate the degree to which signal strength is affected within the first grid. The first signal coverage information is the actual signal coverage information of the access point device within the first grid, and the second signal coverage information is the actual signal coverage information of the access point device within the second grid. The first grid is a grid within the second grid. The acquisition module is further configured to acquire multiple access probabilities and the actual signal coverage information corresponding to each access probability; wherein, the access probability is the access probability of a terminal in the first grid corresponding to the access probability accessing the access point device; The determining module is used to determine the second signal coverage information based on multiple access probabilities and the actual signal coverage information corresponding to each access probability; The training module is used to train the initial signal coverage evaluation model based on the model training data to obtain the target signal coverage evaluation model; the target signal coverage evaluation model is used to predict the signal coverage information of the access point device in the first grid based on the environmental information of the access point device in the first grid and the device information of the access point device.

8. A signal evaluation device, characterized in that, Includes an acquisition module and a processing module; The acquisition module is used to acquire device information of the access point device and environmental information of the access point device within the first grid, wherein the environmental information is used to indicate the degree to which the signal strength is affected within the first grid. The processing module is used to input the device information of the access point device and the environmental information of the access point device in the first grid into the target signal coverage evaluation model to obtain the signal coverage information of the access point device in the first grid. The target signal coverage evaluation model is obtained by training an initial signal coverage evaluation model based on model training parameters. The model training parameters include: a signal evaluation method, environmental information of the access point device within a first grid, device information of the access point device, first signal coverage information, and second signal coverage information. The environmental information indicates the degree to which signal strength is affected within the first grid. The first signal coverage information is the actual signal coverage information of the access point device within the first grid, where the first grid is a grid within a second grid. The second signal coverage information is the actual signal coverage information of the access point device within the second grid. The second signal coverage information indicates the average value of the actual signal coverage information of the access point device in multiple first grids within the second grid. The second signal coverage information is determined based on multiple access probabilities and the actual signal coverage information corresponding to each access probability. The access probability is the probability that a terminal within the first grid corresponding to the access probability will access the access point device.

9. An electronic device, characterized in that, The electronic device includes: processor; A memory configured to store processor-executable instructions; The processor is configured to execute the instructions to implement the signal evaluation method as described in claim 1, or the model training method as described in any one of claims 2-6.

10. A computer-readable storage medium storing instructions thereon, characterized in that, When the instructions in the computer-readable storage medium are executed by an electronic device, the electronic device is able to perform the signal evaluation method as described in claim 1, or the model training method as described in any one of claims 2-6.

Images