Network index prediction method and device, electronic equipment and storage medium

Abstract

The embodiment of the application provides a network index prediction method and device, electronic equipment and a storage medium, and belongs to the fields of artificial intelligence and network communication technology. The method comprises the following steps: obtaining sample prediction data of a network index to be predicted; inputting first sample network data into a target data prediction model to obtain first prediction network data; obtaining prediction error data according to the first prediction network data and second sample network data, and determining a prediction result according to the prediction error data; if the prediction result indicates that the prediction error data meets a preset prediction condition, obtaining first target network data, inputting the first target network data into the target data prediction model to obtain second target network data; and inputting the first target network data and the second target network data into a target network index prediction model for network index prediction. The embodiment of the application can use data that is not easy to obtain to predict a network index, thereby improving the accuracy of prediction.

Description

Technical Field

[0001] This application relates to the fields of artificial intelligence technology and network communication technology, and in particular to a method and apparatus for predicting network indicators, an electronic device and a storage medium. Background Technology

[0002] User experience rate (UXR) refers to the rate at which a user receives data per unit of time, representing the transmission rate achievable by a user in a real network environment. Therefore, UXR can serve as a performance indicator for network optimization. Related technologies use machine learning methods such as deep neural networks to predict UXR. Specifically, a prediction model is built, and data collected from the real network environment is used as input; the output of the prediction model is the UXR. However, in the above methods, the input data may contain data that is difficult to obtain, referring to data requiring complex calculations or prediction operations. For this difficult-to-obtain data, related technologies employ two approaches: First, treating this data as known data and directly inputting it into the prediction model. The advantage of this is that the prediction model can output good prediction results, but because this data is difficult to obtain, it presents practical difficulties. Second, discarding this data reduces the amount of input data, leading to a decrease in the prediction accuracy of the model. Therefore, how to provide a method that can utilize this difficult-to-obtain data for prediction while ensuring prediction accuracy has become an urgent technical problem to be solved. Summary of the Invention

[0003] The main objective of this application is to provide a network indicator prediction method, apparatus, electronic device, and storage medium, which aims to improve the accuracy of prediction by utilizing data that is not easily obtained.

[0004] To achieve the above objectives, a first aspect of this application proposes a network metric prediction method, the method comprising:

[0005] Obtain sample prediction data for the network metric to be predicted; wherein, the sample prediction data includes first sample network data and second sample network data, and the first sample network data is easier to obtain than the second sample network data;

[0006] The first sample network data is input into a preset target data prediction model to predict the network data, thereby obtaining the first predicted network data.

[0007] Prediction error data is obtained based on the first prediction network data and the second sample network data, and the prediction result is determined based on the prediction error data.

[0008] If the prediction result indicates that the prediction error data meets the preset prediction conditions, the first target network data of the network index to be predicted is obtained, and the first target network data is input into the target data prediction model to perform network data prediction to obtain the second target network data; wherein, the data type of the first target network data is consistent with the data type of the first sample network data.

[0009] The first target network data and the second target network data are input into a preset target network indicator prediction model to predict network indicators.

[0010] In some embodiments, before inputting the first target network data and the second target network data into a preset target network indicator prediction model for network indicator prediction, the method further includes training the target network indicator prediction model, specifically including:

[0011] If the prediction result indicates that the prediction error data meets the prediction condition, the first sample network data and the first prediction network data are input into the preset original network index prediction model to perform network index prediction and obtain the prediction index data.

[0012] Obtain sample index data of the network index to be predicted;

[0013] The original network indicator prediction model is adjusted based on the predicted indicator data and the sample indicator data to obtain the target network indicator prediction model.

[0014] In some embodiments, before inputting the first sample network data into a preset target data prediction model to perform network data prediction and obtain the first predicted network data, the method further includes training the target data prediction model, specifically including:

[0015] The first sample network data is input into a preset raw data prediction model to predict the network data, thereby obtaining the second predicted network data.

[0016] The parameters of the original data prediction model are adjusted based on the second sample network data and the second prediction network data to obtain the target data prediction model.

[0017] In some embodiments, the network metric to be predicted includes user experience rate, and obtaining sample metric data of the network metric to be predicted includes:

[0018] Obtain the total downlink data traffic, the data traffic at predetermined time intervals, and the data transmission duration;

[0019] The traffic difference data is calculated based on the total downlink data traffic and the interval data traffic.

[0020] The sample index data is calculated based on the traffic difference data and the data transmission duration.

[0021] In some embodiments, obtaining sample prediction data for the network metric to be predicted includes:

[0022] Obtain raw prediction data, which is data collected by a preset base station;

[0023] Determine the amount of change in the original prediction data within a preset time period, and obtain preliminary prediction data from the original prediction data based on the amount of change;

[0024] The sample prediction data is obtained by filtering from the preliminary prediction data and the preset prediction data.

[0025] To achieve the above objectives, a second aspect of this application provides a network metric prediction apparatus, the apparatus comprising:

[0026] The data acquisition module is used to acquire sample prediction data of the network metric to be predicted; wherein, the sample prediction data includes first sample network data and second sample network data, and the first sample network data is easier to acquire than the second sample network data.

[0027] The first network data prediction model is used to input the first sample network data into a preset target data prediction model to perform network data prediction and obtain the first predicted network data.

[0028] The prediction result determination module is used to obtain prediction error data based on the first prediction network data and the second sample network data, and to determine the prediction result based on the prediction error data.

[0029] The second network data prediction module is used to obtain the first target network data of the network index to be predicted if the prediction result indicates that the prediction error data meets the preset prediction conditions, and input the first target network data into the target data prediction model to perform network data prediction to obtain the second target network data; wherein, the data type of the first target network data is consistent with the data type of the first sample network data.

[0030] The network indicator prediction module is used to input the first target network data and the second target network data into a preset target network indicator prediction model to predict network indicators.

[0031] To achieve the above objectives, a third aspect of the present application provides an electronic device, the electronic device including a memory and a processor, the memory storing a computer program, and the processor executing the computer program to implement the method described in the first aspect.

[0032] To achieve the above objectives, a fourth aspect of the present application provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the method described in the first aspect.

[0033] The network indicator prediction method, apparatus, electronic device, and storage medium proposed in this application obtain second target network data based on the target data prediction model when the prediction result indicates that the prediction error data meets the prediction conditions. This achieves prediction of second target network data based on first target network data, avoiding the difficulty of obtaining second target network data in related technologies. Then, the network indicator to be predicted is predicted based on the first and second target network data, avoiding the low prediction accuracy problem when using only first target network data for network indicator prediction in related technologies. Attached Figure Description

[0034] Figure 1 This is a flowchart of the network metric prediction method provided in the embodiments of this application;

[0035] Figures 2A to 2C This is a schematic diagram of the example sample prediction data of this application;

[0036] Figure 3 yes Figure 1 The flowchart of step S101 in the text;

[0037] Figure 4 Is Figure 1 Before step S102, here is another flowchart of the network metric prediction method provided in this application example;

[0038] Figure 5 yes Figure 1 Before step S105, here is another flowchart of the network metric prediction method provided in this application example;

[0039] Figure 6 yes Figure 5 The flowchart of step S502 in the document;

[0040] Figures 7A to 7B This is a schematic diagram of the prediction error during model training in an embodiment of this application;

[0041] Figure 8This is a schematic diagram of the network index prediction device provided in the embodiments of this application;

[0042] Figure 9 This is a schematic diagram of the hardware structure of the electronic device provided in the embodiments of this application. Detailed Implementation

[0043] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0044] It should be noted that although functional modules are divided in the device schematic diagram and a logical order is shown in the flowchart, in some cases, the steps shown or described may be performed in a different order than the module division in the device or the order in the flowchart. The terms "first," "second," etc., in the specification, claims, and the aforementioned drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence.

[0045] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used herein is for the purpose of describing embodiments of this application only and is not intended to limit this application.

[0046] First, let's analyze some of the terms used in this application:

[0047] Artificial intelligence (AI) is a new branch of computer science that studies, develops, and applies theories, methods, technologies, and systems to simulate, extend, and expand human intelligence. It aims to understand the essence of intelligence and produce intelligent machines that can react in a way similar to human intelligence. Research in this field includes robotics, speech recognition, image recognition, natural language processing, and expert systems. AI can simulate the information processes of human consciousness and thought. Furthermore, AI utilizes digital computers or machines controlled by digital computers to simulate, extend, and expand human intelligence, perceiving the environment, acquiring knowledge, and using that knowledge to achieve optimal results.

[0048] User experience rate (UXR) refers to the rate at which a user receives data per unit of time, representing the transmission rate achievable by a user in a real network environment. Therefore, UXR can serve as a performance indicator for network optimization. Related technologies use machine learning methods such as deep neural networks to predict UXR. Specifically, a prediction model is built, and data collected from the real network environment (including physical layer measurements, user scheduling information, user throughput, etc.) is used as input. The output of the prediction model is the UXR. However, in the above methods, the input data may contain data that is difficult to obtain, referring to data requiring complex calculations or prediction operations. For this difficult-to-obtain data, related technologies employ two approaches: First, directly inputting this data as known data into the prediction model. This approach allows the prediction model to output better prediction results, but it presents practical difficulties due to the difficulty in obtaining this data. Second, discarding this data reduces the amount of input data, leading to a decrease in the prediction accuracy of the model. Therefore, providing a method that can utilize this difficult-to-obtain data for prediction while ensuring accuracy has become a pressing technical problem.

[0049] Based on this, embodiments of this application provide a network metric prediction method and apparatus, electronic device and storage medium, which aim to enable network metric prediction using data that is not easily obtained.

[0050] The network metric prediction method, apparatus, electronic device, and storage medium provided in this application are specifically described through the following embodiments. First, the network metric prediction in the embodiments of this application is described.

[0051] The embodiments of this application can acquire and process relevant data based on artificial intelligence technology. Artificial intelligence (AI) refers to the theories, methods, technologies, and application systems that use digital computers or machines controlled by digital computers to simulate, extend, and expand human intelligence, perceive the environment, acquire knowledge, and use that knowledge to obtain optimal results.

[0052] Foundational technologies for artificial intelligence generally include sensors, dedicated AI chips, cloud computing, distributed storage, big data processing, operating / interactive systems, and mechatronics. AI software technologies mainly encompass computer vision, robotics, biometrics, speech processing, natural language processing, and machine learning / deep learning.

[0053] The network metric prediction method provided in this application relates to the field of artificial intelligence technology. This method can be applied to a terminal, a server, or software running on either a terminal or a server. In some embodiments, the terminal can be a smartphone, tablet, laptop, desktop computer, etc.; the server can be configured as an independent physical server, a server cluster or distributed system composed of multiple physical servers, or a cloud server providing basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, CDN, and big data and artificial intelligence platforms; the software can be an application implementing the network metric prediction method, but is not limited to the above forms.

[0054] This application can be used in a wide variety of general-purpose or special-purpose computer system environments or configurations. Examples include: personal computers, server computers, handheld or portable devices, tablet devices, multiprocessor systems, microprocessor-based systems, set-top boxes, programmable consumer electronics, network PCs, minicomputers, mainframe computers, and distributed computing environments including any of the above systems or devices. This application can be described in the general context of computer-executable instructions executed by a computer, such as program modules. Generally, program modules include routines, programs, objects, components, data structures, etc., that perform specific tasks or implement specific abstract data types. This application can also be practiced in distributed computing environments where tasks are performed by remote processing devices connected via a communication network. In distributed computing environments, program modules can reside in local and remote computer storage media, including storage devices.

[0055] Figure 1 This is an optional flowchart of the network metric prediction method provided in the embodiments of this application. Figure 1 The method may include, but is not limited to, steps S101 to S105.

[0056] Step S101: Obtain sample prediction data of the network metric to be predicted; wherein, the sample prediction data includes first sample network data and second sample network data, and the first sample network data is easier to obtain than the second sample network data.

[0057] Step S102: Input the first sample network data into the preset target data prediction model to predict the network data and obtain the first predicted network data;

[0058] Step S103: Obtain prediction error data based on the first prediction network data and the second sample network data, and determine the prediction result based on the prediction error data;

[0059] Step S104: If the prediction result indicates that the prediction error data meets the preset prediction conditions, obtain the first target network data of the network index to be predicted, input the first target network data into the target data prediction model to predict the network data, and obtain the second target network data; wherein, the data type of the first target network data is consistent with the data type of the first sample network data.

[0060] Step S105: Input the first target network data and the second target network data into the preset target network index prediction model to predict the network index.

[0061] In steps S101 to S105 of this embodiment, when the prediction result indicates that the prediction error data meets the prediction conditions, the second target network data is obtained according to the target data prediction model. This achieves the prediction of the second target network data based on the first target network data, avoiding the problem of difficulty in obtaining the second target network data in related technologies. Then, the network index to be predicted is predicted based on the first and second target network data, avoiding the problem of low prediction accuracy when using only the first target network data for network index prediction in related technologies.

[0062] In step S101 of some embodiments, the network metric to be predicted refers to an indicator that needs to be predicted and can represent the network communication performance. For example, it can be any one of user experience rate, latency, area transmission capacity, peak data rate, etc. This application does not specifically limit this. However, for ease of explanation, in the following embodiments, the user experience rate is used as the network metric to be predicted. Sample prediction data refers to data related to the network metric to be predicted, that is, the network metric to be predicted can be predicted based on the sample prediction data. Sample prediction data can be obtained from base stations, terminals used by users, etc. Figures 2A to 2C The image shows sample prediction data obtained when the network metric to be predicted is user experience rate. It is understandable that... Figures 2A to 2C The specific numerical values ​​of the predicted sample data are not shown; instead, each predicted sample data point is represented by an identifier character. This identifier character indicates the data type of the corresponding sample data. In other words, in... Figures 2A to 2C In this process, 74-dimensional sample prediction data was obtained. However, it should be understood that although... Figures 2A to 2C The example illustrates 74-dimensional sample prediction data, but this application does not specifically limit the dimensionality of the acquired sample prediction data. The 74-dimensional sample prediction data can be categorized into the following seven types:

[0063] The first category, including data from dimensions 1 to 11, is data related to RSRP (Reference Signal Received Power).

[0064] The second category, including data from dimensions 12 to 27, consists of data related to antennas.

[0065] The third category, including data from dimensions 28 to 33, consists of data related to scheduling.

[0066] The fourth category, including dimensions 35 to 38, consists of data related to PRBs (Physical Resource Blocks).

[0067] The fifth category, including data from dimensions 39 to 52, consists of data related to MU (Multi-User) and MIMO (Multiple-Input Multiple-Output).

[0068] The sixth category, including dimensions 53 and 64, consists of data related to time statistics.

[0069] The seventh category, including dimensions 65 to 74, consists of data related to scheduling.

[0070] Among them, the third and seventh categories of data are both scheduling-related data. The difference lies in the ease of obtaining the third category, which is greater than that of the seventh category. In other words, the seventh category of data is not easy to obtain and requires complex calculations and multi-step predictions. By analogy, the easily obtainable data described in the embodiments of this application refers to data that can be directly collected from the base station or obtained through simple calculations and simple prediction steps. Therefore, the sample prediction data includes the first sample network data and the second sample network data. The first sample network data is easily obtainable (such as the first to fourth and sixth categories of data, totaling 50 dimensions), while the second sample network data is not easily obtainable (such as the fifth and seventh categories of data, totaling 24 dimensions).

[0071] Reference Figure 3 In some embodiments, step S101 includes, but is not limited to, steps S301 to S303.

[0072] Step S301: Obtain the raw prediction data, which is the data collected by the preset base station;

[0073] Step S302: Determine the amount of change in the original prediction data within a preset time period, and obtain preliminary prediction data from the original prediction data based on the amount of change;

[0074] Step S303: Select sample prediction data from the preliminary prediction data and the preset prediction data.

[0075] In step S301 of some embodiments, raw prediction data collected by the base station is acquired. This raw prediction data is data related to network communication. Based on different collection times and the identity information of different collection objects, multiple raw prediction data are stored in different event block measurement entries.

[0076] In step S302 of some embodiments, the change amount refers to the numerical change of the original prediction data within a preset time period. If the change amount is zero, it indicates that the original prediction data has not changed. Since the values ​​corresponding to the network indicators to be predicted change, the original prediction data with zero change amount is considered to have a weak correlation with the network indicators to be predicted. Therefore, the original prediction data can be filtered based on the change amount, removing the original prediction data with zero change amount, and using the remaining original prediction data as the preliminary prediction data.

[0077] In step S303 of some embodiments, in addition to the data collected by the base station, data related to the network indicators to be predicted that can be easily obtained through other means is also included, utilizing data obtained from the user terminal. This portion of data is used as preset prediction data. The preset prediction data is integrated with the preliminary prediction data, and sample prediction data is selected from the integrated dataset. The integrated dataset can be selected based on factors such as the strength of its relevance to the event to be predicted, historical experience, and the ease of data acquisition; this embodiment does not specifically limit this. It is understood that when obtaining relevant data from the user terminal, user authorization is obtained in advance via a pop-up window.

[0078] Reference Figure 4 In some embodiments, before step S102, the method provided in this application embodiment further includes the step of training a target data prediction model, specifically including but not limited to steps S401 to S402.

[0079] Step S401: Input the first sample network data into the preset original data prediction model to predict the network data and obtain the second predicted network data;

[0080] Step S402: Adjust the parameters of the original data prediction model based on the second sample network data and the second prediction network data to obtain the target data prediction model.

[0081] In step S401 of some embodiments, a raw data prediction model including an input layer, intermediate layers, and an output layer is pre-set. A BatchNorm layer can be added to the input side of each network layer, with the BatchSize set to 64. A Dropout layer can also be added, with a dropout probability set to 0.2. The activation function is ReLU (Rectified Linear Function). All first sample network data (e.g., 50-dimensional data) are used as input data to the raw data prediction model to obtain the second prediction network data predicted by the raw data prediction model. The second prediction network data is data of the same type as the second sample network data.

[0082] In step S402 of some embodiments, the prediction error of the original data prediction model is calculated based on the second sample network data and the second prediction network data. The parameters of the original data prediction model are then adjusted based on the prediction error until the original data prediction model converges, thus obtaining the target data prediction model. Convergence conditions include the number of iterations reaching a preset number (e.g., 20 times) and the loss function decreasing by no more than a preset loss function threshold.

[0083] The advantage of steps S401 to S402 is that the hidden relationship between the first sample network data and the second sample network data is mined through the target data prediction model. This allows the second sample network data to be predicted based on the first sample network data, thus enabling the acquisition of the second sample network data even when only the first sample network data is needed, without requiring the acquisition of the second sample network data. Furthermore, this embodiment of the application performs direct prediction based on the target data prediction model, avoiding the need for complex calculations to obtain the second sample network data in related technologies.

[0084] In step S102 of some embodiments, the target data prediction model is a pre-trained model capable of predicting data of a preset type. The preset type refers to a type consistent with the data type of the second sample network data. Therefore, when all the first sample network data is input into the target data prediction model, it can predict first prediction network data of the preset type. In other words, the first prediction network data is the prediction result of the target data prediction model for the second sample network data.

[0085] In step S103 of some embodiments, the prediction error data refers to the prediction error of the target data prediction model for the second sample network data. That is, although the target data prediction model is a trained model, inaccurate predictions may still occur when the correlation between the first and second sample network data is weak. In this case, it is necessary to use the prediction error data to determine which data types of the second sample network data the target data prediction model can accurately predict. Specifically, when the prediction error data is less than or equal to a preset threshold, it is determined that the target data prediction model predicts the corresponding data type relatively accurately; when the prediction error data is greater than the preset threshold, it is determined that the target data prediction model predicts the corresponding data type inaccurately.

[0086] In step S104 of some embodiments, it is assumed that the prediction condition is that the prediction error data is less than or equal to a preset threshold. According to step S103, the prediction result when the prediction error data is less than (or greater than, or equal to) the preset threshold can be obtained. When the prediction result indicates that the prediction error data meets the prediction condition, it indicates that the target data prediction model has high prediction accuracy for the corresponding data type. At this time, the prediction result of the target data prediction model can be used to replace the corresponding second sample network data. Table 1 below shows the prediction results of the target data prediction model for the fifth type of data.

[0087] AccumMeasItems.DLMuSchCnt 29.76 AccumMeasItems.DLMuSchRankRbNum 39.76 AccumMeasItems.DLMuSchRbNum 39.07 AccumMeasItems.DLMuSchMCSCode0 50.01 AccumMeasItems.DLMuSchMCSCntCode0 29.79 AccumMeasItems.DLMuBeamPairCnt 29.47 AccumMeasItems.DLMuBeamPairRankRbNum 40.24 MIMOAveDLSchRankRb 5.51 MIMOAveDLSchRb 1.98 MIMOAveMCS 8.94 MIMOAveDLMuSchRankRb 12.58 MIMOAveDLMuSchRb 12.63 MIMOAveMuMCS 34.71 MIMOAveBeamPair 12.60

[0088] Table 1

[0089] Table 2 below shows the results of the target data prediction model's prediction of the seventh type of data.

[0090] AveDLRank 3.46 AveDLMCS 14.25 UEMeasInfo.AccumMeasItems.DLRank 4.99 UEMeasInfo.AccumMeasItems.DlMCSCode0 16.75 UEMeasInfo.AccumMeasItems.DlMcsCntCode0 0.38 AccumMeasItems.DLTotalSchCnt 0.34 AccumMeasItems.DLTotalSchRankRbNum 7.39 AccumMeasItems.DLTotalSchRbNum 2.30 AccumMeasItems.DLInitMcsCode0 11.43 AccumMeasItems.DLInitMcsCntCode0 0.34

[0091] Table 2

[0092] Assuming a preset threshold of 5%, Tables 1 and 2 show that when the target data prediction network makes predictions based on all first sample network data, it can accurately predict the following seven types of second sample network data: MIMOAveDLSchRb, AveDLRank, UEMeasInfo.AccumMeasItems.DLRank, UEMeasInfo.AccumMeasItems.DlMcsCntCode0, AccumMeasItems.DLTotalSchCnt, AccumMeasItems.DLTotalSchRbNum, and AccumMeasItems.DLInitMcsCntCode0.

[0093] Therefore, in practical applications, when the acquired first target network data (with the same data type as the first sample network data) is used as the prediction data for the target data prediction model, a second target network data with a small error compared to the actual data (i.e., data with the same data type as the second sample network data in the actual application stage) will be obtained. In this case, the second target network data can be used to replace the actual data, thus avoiding the difficulty in obtaining actual data in related technologies. Taking Tables 1 and 2 as examples, the target data prediction model will output 7-dimensional second target network data.

[0094] It is understood that the prediction error data in the embodiments of this application refers to the absolute percentage value of the first prediction network data and the second sample network data.

[0095] Reference Figure 5In some embodiments, before step S105, the method provided in this application embodiment further includes the step of training a target network index prediction model, specifically including but not limited to steps S501 to S503.

[0096] Step S501: If the prediction result indicates that the prediction error data meets the prediction conditions, input the first sample network data and the first prediction network data into the preset original network index prediction model to perform network index prediction and obtain the prediction index data.

[0097] Step S502: Obtain sample index data of the network index to be predicted;

[0098] Step S503: Adjust the parameters of the original network indicator prediction model based on the predicted indicator data and sample indicator data to obtain the target network indicator prediction model.

[0099] In step S501 of some embodiments, the original network indicator prediction model is a pre-set model with the same or different model structure as the original data prediction model. If the prediction result indicates that the prediction error data meets the prediction conditions, it means that the first prediction network data can be used to replace the corresponding second sample network data. At this time, the first prediction network data and the first sample network data can be used as input data for the original network indicator prediction model to predict the network indicator to be predicted and obtain the predicted indicator data.

[0100] In step S502 of some embodiments, the numerical value corresponding to the network indicator data to be predicted obtained from the real network environment is acquired, i.e., the sample indicator data. It can be understood that the sample indicator data is the data corresponding to the first sample network data and the second sample network data, that is, the first sample network data and the second sample network data determine the size of the sample indicator data.

[0101] Reference Figure 6 In some embodiments, step S502 includes, but is not limited to, steps S601 to S603.

[0102] Step S601: Obtain the total downlink data traffic, the interval data traffic for a predetermined time interval, and the data transmission duration;

[0103] Step S602: Calculate the traffic difference data based on the total downlink data traffic and the interval data traffic;

[0104] Step S603: Calculate sample indicator data based on traffic difference data and data transmission duration.

[0105] It should be noted that, since downlink IP throughput is an important performance metric for user experience rate, in this embodiment, downlink IP throughput (IP_Thp) represents user experience rate. That is, steps S601 to S603 are the calculation steps for downlink IP throughput, which will be explained in detail below.

[0106] In steps S601 to S603 of some embodiments, the downlink total data traffic (ThpVol), the interval data traffic (ThpVolLastSlot) of a predetermined time interval, and the data transmission duration (ThpTimeRmvLastSlot) of the RLC (Radio Link Control) layer are obtained respectively, and then the downlink IP throughput is calculated according to the following formula (1).

[0107]

[0108] It is understandable that the predetermined time interval is the transmission time interval, and the interval data traffic refers to the data traffic corresponding to the last transmission time interval within a certain duration. The data transmission duration refers to the remaining data transmission duration after removing the last transmission time interval from the aforementioned certain duration.

[0109] In step S503 of some embodiments, the prediction error of the original network indicator prediction model is calculated based on a preset loss function, prediction indicator data, and sample indicator data. The parameters of the original network indicator prediction model are then adjusted based on the prediction error until the original network indicator prediction model converges, thus obtaining the target network indicator prediction model. Convergence conditions include the number of iterations reaching a preset number (e.g., 20 times) and the loss function decreasing by no more than a preset loss function threshold.

[0110] It is understandable that the prediction error MAPE of the original data prediction model can be calculated based on the loss function shown in Equation (2).

[0111]

[0112] Where N represents the number of samples in the sample indicator data, y k This represents the predictive indicator data. This represents the sample index data. It is understandable that when training the prediction model for the target data, the loss function shown in equation (2) can also be selected.

[0113] The advantage of steps S501 to S503 is that the hidden relationship between the first sample network data, the first predicted network data and the network indicator to be predicted is mined through the target network indicator prediction model, so that the network indicator to be predicted can be predicted based on the first sample network data and the first predicted network data.

[0114] In step S105 of some embodiments, the target network metric prediction model is a pre-trained model with the ability to predict the network metric to be predicted, such as the ability to predict user experience rate. The first target network data and the second target network data are used as input data to the target network metric prediction model, and the target network metric to be predicted in the actual application is predicted according to the target network metric prediction model.

[0115] The feasibility of this application example is explained below based on experimental data. The sample prediction data is divided into a training set and a test set according to a certain ratio. (Refer to...) Figure 7A This diagram illustrates the prediction error when only the first sample network data is used for user experience rate prediction, excluding the second sample network data. (Refer to...) Figure 7B This diagram illustrates the prediction error when predicting user experience rate based on first sample network data (50-dimensional data) and first prediction network data (e.g., 7-dimensional data) that meets the prediction condition (prediction error less than or equal to 5%). Figure 7A and Figure 7B A comparison shows that the method provided in the embodiments of this application ( Figure 7B The corresponding method has a better prediction effect on user experience rate, indicating that the method provided in the embodiments of this application has a certain feasibility.

[0116] Reference Figure 8 This application also provides a network metric prediction apparatus that can implement the above-described network metric prediction method. The apparatus includes:

[0117] The data acquisition module 801 is used to acquire sample prediction data of the network indicators to be predicted; wherein, the sample prediction data includes first sample network data and second sample network data, and the first sample network data is easier to acquire than the second sample network data.

[0118] The first network data prediction model 802 is used to input the first sample network data into the preset target data prediction model to perform network data prediction and obtain the first predicted network data.

[0119] The prediction result determination module 803 is used to obtain prediction error data based on the first prediction network data and the second sample network data, and to determine the prediction result based on the prediction error data.

[0120] The second network data prediction module 804 is used to obtain the first target network data of the network index to be predicted if the prediction result indicates that the prediction error data meets the preset prediction conditions, and input the first target network data into the target data prediction model to perform network data prediction to obtain the second target network data; wherein, the data type of the first target network data is consistent with the data type of the first sample network data.

[0121] The network indicator prediction module 805 is used to input the first target network data and the second target network data into the preset target network indicator prediction model to predict the network indicators.

[0122] The specific implementation of this network indicator prediction device is basically the same as the specific implementation of the network indicator prediction method described above, and will not be repeated here.

[0123] This application also provides an electronic device, which includes a memory and a processor. The memory stores a computer program, and the processor executes the computer program to implement the aforementioned network metric prediction method. This electronic device can be any smart terminal, including tablet computers, in-vehicle computers, etc.

[0124] Please see Figure 9 , Figure 9 The hardware structure of an electronic device according to another embodiment is illustrated. The electronic device includes:

[0125] The processor 901 can be implemented using a general-purpose CPU (Central Processing Unit), microprocessor, application-specific integrated circuit (ASIC), or one or more integrated circuits, and is used to execute relevant programs to implement the technical solutions provided in the embodiments of this application.

[0126] The memory 902 can be implemented as a read-only memory (ROM), a static storage device, a dynamic storage device, or a random access memory (RAM). The memory 902 can store the operating system and other applications. When the technical solutions provided in the embodiments of this specification are implemented through software or firmware, the relevant program code is stored in the memory 902 and is called and executed by the processor 901 using the network metric prediction method of the embodiments of this application.

[0127] The input / output interface 903 is used to implement information input and output;

[0128] The communication interface 904 is used to enable communication and interaction between this device and other devices. Communication can be achieved through wired means (such as USB, Ethernet cable, etc.) or wireless means (such as mobile network, WIFI, Bluetooth, etc.).

[0129] Bus 905 transmits information between various components of the device (e.g., processor 901, memory 902, input / output interface 903, and communication interface 904);

[0130] The processor 901, memory 902, input / output interface 903, and communication interface 904 are connected to each other within the device via bus 905.

[0131] This application also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described network metric prediction method.

[0132] Memory, as a non-transitory computer-readable storage medium, can be used to store non-transitory software programs and non-transitory computer-executable programs. Furthermore, memory may include high-speed random access memory, and may also include non-transitory memory, such as at least one disk storage device, flash memory device, or other non-transitory solid-state storage device. In some embodiments, memory may optionally include memory remotely located relative to the processor, and these remote memories can be connected to the processor via a network. Examples of such networks include, but are not limited to, the Internet, intranets, local area networks, mobile communication networks, and combinations thereof.

[0133] The embodiments described in this application are for the purpose of more clearly illustrating the technical solutions of the embodiments of this application, and do not constitute a limitation on the technical solutions provided by the embodiments of this application. As those skilled in the art will know, with the evolution of technology and the emergence of new application scenarios, the technical solutions provided by the embodiments of this application are also applicable to similar technical problems.

[0134] Those skilled in the art will understand that the technical solutions shown in the figures do not constitute a limitation on the embodiments of this application, and may include more or fewer steps than shown, or combine certain steps, or different steps.

[0135] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

[0136] Those skilled in the art will understand that all or some of the steps in the methods disclosed above, as well as the functional modules / units in the systems and devices, can be implemented as software, firmware, hardware, or suitable combinations thereof.

[0137] The terms “first,” “second,” “third,” “fourth,” etc. (if present) in the specification and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms “comprising” and “having,” and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0138] It should be understood that in this application, "at least one (item)" means one or more, and "more than" means two or more. "And / or" is used to describe the relationship between related objects, indicating that three relationships can exist. For example, "A and / or B" can represent three cases: only A exists, only B exists, and both A and B exist simultaneously, where A and B can be singular or plural. The character " / " generally indicates that the preceding and following related objects are in an "or" relationship. "At least one (item) of the following" or similar expressions refer to any combination of these items, including any combination of single or plural items. For example, at least one (item) of a, b, or c can represent: a, b, c, "a and b", "a and c", "b and c", or "a and b and c", where a, b, and c can be single or multiple.

[0139] In the several embodiments provided in this application, it should be understood that the disclosed apparatus and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of the units described above is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or units may be electrical, mechanical, or other forms.

[0140] The units described above 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.

[0141] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0142] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes multiple instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this application. The aforementioned storage medium includes various media capable of storing programs, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0143] The preferred embodiments of the present application have been described above with reference to the accompanying drawings, but this does not limit the scope of the claims of the present application. Any modifications, equivalent substitutions, and improvements made by those skilled in the art without departing from the scope and substance of the embodiments of the present application shall be within the scope of the claims of the present application.

Claims

1. A method for predicting network indicators, characterized in that, The method includes: Obtain sample prediction data for the network metric to be predicted; wherein, the sample prediction data includes first sample network data and second sample network data, and the first sample network data is easier to obtain than the second sample network data; The first sample network data is input into a preset target data prediction model to predict the network data, thereby obtaining the first predicted network data. Prediction error data is obtained based on the first prediction network data and the second sample network data, and the prediction result is determined based on the prediction error data. If the prediction result indicates that the prediction error data meets the preset prediction conditions, the first target network data of the network index to be predicted is obtained, and the first target network data is input into the target data prediction model to perform network data prediction to obtain the second target network data; wherein, the data type of the first target network data is consistent with the data type of the first sample network data. The first target network data and the second target network data are input into a preset target network indicator prediction model to predict network indicators. The step of obtaining sample prediction data for the network metric to be predicted includes: Obtain raw prediction data, which is data collected by a preset base station; Determine the amount of change in the original prediction data within a preset time period, and obtain preliminary prediction data from the original prediction data based on the amount of change; The sample prediction data is obtained by filtering from the preliminary prediction data and the preset prediction data. The preset prediction data is data obtained by user terminals that are related to the network indicators to be tested, other than the data collected by the base station.

2. The method according to claim 1, characterized in that, Before inputting the first target network data and the second target network data into a preset target network indicator prediction model for network indicator prediction, the method further includes training the target network indicator prediction model, specifically including: If the prediction result indicates that the prediction error data meets the prediction condition, the first sample network data and the first prediction network data are input into the preset original network index prediction model to perform network index prediction and obtain the prediction index data. Obtain sample index data of the network index to be predicted; The original network indicator prediction model is adjusted based on the predicted indicator data and the sample indicator data to obtain the target network indicator prediction model.

3. The method according to claim 2, characterized in that, The network metric to be predicted includes user experience rate, and obtaining sample metric data for the network metric to be predicted includes: Obtain the total downlink data traffic, the data traffic at predetermined time intervals, and the data transmission duration; The traffic difference data is calculated based on the total downlink data traffic and the interval data traffic. The sample index data is calculated based on the traffic difference data and the data transmission duration.

4. A network indicator prediction device, characterized in that, The device includes: The data acquisition module is used to acquire sample prediction data of the network metric to be predicted; wherein, the sample prediction data includes first sample network data and second sample network data, and the first sample network data is easier to acquire than the second sample network data. The first network data prediction module is used to input the first sample network data into a preset target data prediction model to perform network data prediction and obtain the first predicted network data. The prediction result determination module is used to obtain prediction error data based on the first prediction network data and the second sample network data, and to determine the prediction result based on the prediction error data. The second network data prediction module is used to obtain the first target network data of the network index to be predicted if the prediction result indicates that the prediction error data meets the preset prediction conditions, and input the first target network data into the target data prediction model to perform network data prediction to obtain the second target network data; wherein, the data type of the first target network data is consistent with the data type of the first sample network data. The network indicator prediction module is used to input the first target network data and the second target network data into a preset target network indicator prediction model to predict network indicators. The step of obtaining sample prediction data for the network metric to be predicted includes: Obtain raw prediction data, which is data collected by a preset base station; Determine the amount of change in the original prediction data within a preset time period, and obtain preliminary prediction data from the original prediction data based on the amount of change; The sample prediction data is obtained by filtering from the preliminary prediction data and the preset prediction data. The preset prediction data is data obtained by user terminals that are related to the network indicators to be tested, other than the data collected by the base station.

5. An electronic device, characterized in that, The electronic device includes a memory and a processor, the memory storing a computer program, and the processor executing the computer program to implement the method according to any one of claims 1 to 3.

6. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by a processor, it implements the method of any one of claims 1 to 3.

Images