Method for acquiring encrypted short video experience index data and related device

Abstract

According to the method for obtaining the experience index data of the encrypted short video and the related device, the resolution of the encrypted short video can be identified. The method comprises the following steps: after a probe obtains network feature data of a service time period according to an encrypted short video code stream of the service time period, receiving the network feature data of the service time period sent by the probe, and inputting the network feature data of the service time period into an artificial intelligence model, and outputting the resolution type of the encrypted short video in the service time period through the artificial intelligence model.

Description

Technical Field

[0001] The present application relates to the field of communications, and in particular to a method and related device for obtaining encrypted short video experience indicator data. Background Art

[0002] In order to ensure the service quality of short video services, operators need to obtain experience indicator data of short video services, such as the resolution of short videos.

[0003] Currently, there is a method for measuring the resolution of non-encrypted short videos, which is roughly as follows: obtain the bit rate of the non-encrypted video, and determine the resolution of the non-encrypted video based on the corresponding relationship between the bit rate and resolution of the non-encrypted video.

[0004] However, it is currently impossible to directly obtain the bit rate of the encrypted short video from the operator network, so the operator cannot determine the resolution of the encrypted short video based on the bit rate. Summary of the invention

[0005] The present application provides a method and related device for obtaining encrypted short video experience indicator data, which can accurately identify the resolution of encrypted short videos.

[0006] The first aspect provides a method for obtaining encrypted short video experience indicator data, the method comprising: after the probe obtains network feature data of the business period according to the encrypted short video code stream of the business period, receiving the network feature data of the business period sent by the probe, inputting the network feature data of the business period into an artificial intelligence model, and outputting the resolution type of the encrypted short video within the business period through the artificial intelligence model.

[0007] Among them, the encrypted short video experience index data includes one or more of the short video resolution, the proportion of low-definition types and the proportion of high-definition types. The network feature data includes at least two of the slice data transmission rate, the latency of the data slice, the data volume of the time slice, the maximum continuous low-speed duration and the stall ratio. The probe of the operator's network can obtain the network feature data of the encrypted short video, and the operator's customer experience management device can identify the resolution type of the encrypted short video based on the network feature data of the encrypted short video and the artificial intelligence model, which solves the problem that the operator cannot identify the resolution type of the encrypted short video.

[0008] In a possible implementation, the network feature data of a service period is divided into network feature data of multiple service sub-periods; the network feature data of the service sub-periods is input into an artificial intelligence model, and the resolution type of the encrypted short video in the service sub-period is output through the artificial intelligence model. For each service sub-period, when the resolution type of the encrypted short video in the service sub-period is a low-definition type, the service sub-period is determined to be a low-definition sub-period, and the proportion of the low-definition type is determined according to the total number of low-definition sub-periods and the total number of service sub-periods. In this way, the experience indicator data of the proportion of low-definition types can be obtained.

[0009] In another possible implementation, the network feature data of the service period is divided into the network feature data of multiple service sub-periods; the network feature data of the service sub-periods is input into the artificial intelligence model, and the resolution type of the encrypted short video in the service sub-period is output through the artificial intelligence model. For each service sub-period, when the resolution type of the encrypted short video in the service sub-period is a high-definition type, the service sub-period is determined to be a high-definition sub-period, and the proportion of high-definition types is determined according to the total number of high-definition sub-periods and the total number of service sub-periods. In this way, the experience indicator data of the proportion of high-definition types can be obtained.

[0010] In another possible implementation, the network feature data also includes a cell identifier, and the method for obtaining encrypted short video experience index data of the present application also includes: when the proportion of high-definition types in the service period is less than a first threshold, a cell network construction plan is generated according to the cell corresponding to the cell identifier; when the proportion of low-definition types in the service period is greater than a second threshold, a cell network optimization strategy is generated according to the cell corresponding to the cell identifier. In this way, the cell network strategy can be updated and the cell network quality can be improved.

[0011] In another possible implementation, the method for obtaining encrypted short video experience index data of the present application also includes: receiving multiple encrypted short video feature samples sent by the probe, and training an artificial intelligence model based on the multiple encrypted short video feature samples. The encrypted short video feature samples include network feature data of the encrypted short video and the resolution type of the encrypted short video, and the resolution type of the encrypted short video is a sample label.

[0012] The second aspect provides a method for obtaining short video experience index data, the method comprising: after collecting the encrypted short video code stream sent by the terminal during the business period, determining the network feature data of the business period according to the encrypted short video code stream of the business period, and sending the network feature data of the business period to the customer experience management device. The method can obtain network feature data related to the encrypted short video experience index data. Among them, the network feature data includes at least two of the slice data transmission rate, the delay of the data slice, the data volume of the time slice, the maximum continuous low speed duration and the stall ratio.

[0013] In one possible implementation, determining network characteristic data based on the encrypted short video code stream of the business period includes: removing the idle period from the business period to obtain the effective data transmission duration, removing the retransmitted data packet and the header of each data packet from the encrypted short video code stream of the business period to obtain the effective data; determining the slice data transmission rate according to the amount of effective data and the effective data transmission duration. The idle period is a period greater than the packet interval threshold and no data is transmitted. The header of the data packet does not belong to the encrypted short video data, and the retransmitted data packet will cause the data volume to be calculated repeatedly. No data is sent during the idle period. This implementation can remove the influence of the above factors on the calculation of the transmission rate of the encrypted short video data (i.e., the slice data transmission rate), thereby improving the accuracy of the transmission rate of the encrypted short video data.

[0014] In another possible implementation, determining network characteristic data according to the encrypted short video code stream of the business period includes: determining short video data packets greater than a data volume threshold from the encrypted short video code stream of the business period; dividing the short video data packets of the business period into multiple data slices according to a preset slice size; and determining the slice data transmission rate according to the data volume of all data slices and the duration of all data slices. In this way, short video data packets can be screened out according to the data volume threshold, reducing the impact of signaling data on the slice data transmission rate, and improving the accuracy of the slice data transmission rate.

[0015] In another possible implementation, determining network characteristic data according to the encrypted short video code stream of the business period includes: obtaining all transactions from the encrypted short video code stream of the business period, each transaction including a transaction request, data corresponding to the transaction request, and a confirmation message corresponding to the transaction request; determining the slice data transmission rate according to the transaction data volume of all transactions and the transaction duration of all transactions. Since the transaction does not include signaling for establishing a communication connection, the impact of the above signaling on the slice data transmission rate can be reduced, thereby improving the accuracy of the slice data transmission rate.

[0016] In another possible implementation, determining network characteristic data according to the encrypted short video code stream of the service period includes: determining a short video data packet greater than a data volume threshold from the encrypted short video code stream of the service period, dividing the short video data packet of the service period into a plurality of data slices according to a preset slice size, and determining a delay of the data slice. The delay of the data slice is related to the short video resolution and the short video fluency.

[0017] In another possible implementation, determining network characteristic data according to the encrypted short video code stream of the service period includes: determining short video data packets greater than a data volume threshold from the encrypted short video code stream of the service period, dividing the service period into multiple time slices according to a preset slice delay, and determining the data volume of the time slice. The data volume of the time slice is related to the short video resolution and the short video fluency.

[0018] In another possible implementation, determining network characteristic data based on the encrypted short video code stream of the business period includes: selecting multiple time intervals from the business period; for each time interval, when the average data transmission rate of the time interval is less than the low-speed threshold, determining the time interval as a low-speed time interval; determining the stall ratio as the ratio of the number of low-speed time intervals to the total number of time intervals.

[0019] In another possible implementation, determining network characteristic data based on the encrypted short video code stream of the business time period includes: selecting multiple time intervals from the business time period; for each time interval, when the average data transmission rate of the time interval is less than the low-speed threshold, determining the time interval as a low-speed time interval; taking multiple consecutive low-speed time intervals in the business period as low-speed periods; and determining the maximum continuous low-speed duration as the duration of the maximum low-speed period.

[0020] In another possible implementation, before collecting the encrypted short video code stream sent by the terminal, the method of obtaining encrypted short video experience index data in the present application also includes: collecting the service data sent by the terminal during the dialing test period; determining the network feature data of the dialing test period according to the encrypted short video code stream of the dialing test period; obtaining the dialing test data, the dialing test data includes multiple records, and the records include user identification, dialing test period, service identification and resolution type; matching the network feature data included in the service data and the resolution type included in the dialing test data according to the user identification and service identification; selecting the target network feature data and the target resolution type from the successfully matched network feature data and resolution types according to the same period in the playback period and the dialing test period; generating an encrypted short video feature sample according to the target network feature data and the target resolution type; and sending multiple encrypted short video feature samples to the customer experience management device. Among them, the service data includes the encrypted short video code stream, user identification, service period and service identification.

[0021] The third aspect provides a customer experience management device, which has the function of implementing the method for obtaining encrypted short video experience index data in any one of the embodiments of the first aspect. The function can be implemented by hardware, or by hardware executing corresponding software. The hardware or software includes one or more modules corresponding to the above functions, such as a receiving module, a processing module, an experience index module, a strategy module, and a training module.

[0022] In a fourth aspect, a probe is provided, which has the function of implementing the method for obtaining encrypted short video experience index data in any one of the embodiments of the second aspect. The function can be implemented by hardware, or by hardware executing corresponding software. The hardware or software includes one or more modules corresponding to the above functions, such as a receiving module, a processing module, and a sending module.

[0023] The fifth aspect provides a computing device cluster, which includes at least one computing device, each computing device includes a processor and a memory, and the processor of the at least one computing device is used to execute instructions stored in the memory of the at least one computing device, so that the computing device cluster performs the method of the first aspect.

[0024] The sixth aspect provides a computing device cluster, which includes at least one computing device, each computing device includes a processor and a memory, and the processor of the at least one computing device is used to execute instructions stored in the memory of the at least one computing device, so that the computing device cluster performs the method of the second aspect.

[0025] The seventh aspect provides a computer-readable storage medium, in which instructions are stored. When the computer-readable storage medium is run on a computer, the computer executes the method of the first aspect.

[0026] The eighth aspect provides a computer-readable storage medium, in which instructions are stored. When the computer-readable storage medium is run on a computer, the computer executes the method of the second aspect.

[0027] A ninth aspect provides a computer program product comprising instructions, which, when executed on a computer, causes the computer to execute the method of the first aspect.

[0028] The tenth aspect provides a computer program product comprising instructions, which, when executed on a computer, causes the computer to execute the method of the second aspect. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] Figure 1 A schematic diagram of an application scenario in an embodiment of the present application;

[0030] Figure 2 A schematic diagram of the structure of a customer experience management device in an embodiment of the present application;

[0031] Figure 3 A schematic diagram of the structure of the probe in the embodiment of the present application;

[0032] Figure 4 A schematic diagram of training a deep learning model and model reasoning in an embodiment of the present application;

[0033] Figure 5 A signaling interaction diagram for training an artificial intelligence model based on encrypted short video feature samples in an embodiment of the present application;

[0034] Figure 6 A schematic diagram of a method for obtaining encrypted short video experience index data in an embodiment of the present application;

[0035] Figure 7A schematic diagram of the associated features in the embodiment of the present application;

[0036] Figure 8 A schematic diagram of valid data and valid data duration in an embodiment of the present application;

[0037] Fig. 9 A schematic diagram of data slicing in an embodiment of the present application;

[0038] Fig.10 A schematic diagram of a transaction in an embodiment of the present application;

[0039] Fig.11 Another schematic diagram of data slicing in an embodiment of the present application;

[0040] Fig.12 A schematic diagram of time slicing in an embodiment of the present application;

[0041] Fig.13 A schematic diagram of the time interval and stall ratio in the embodiment of the present application;

[0042] Fig.14 A schematic diagram of the low-speed period and the maximum continuous low-speed duration in the embodiment of the present application;

[0043] Fig.15 Another schematic diagram of the method for obtaining encrypted short video experience index data in an embodiment of the present application;

[0044] Fig.16 A structural diagram of a computing device in an embodiment of the present application;

[0045] Fig.17 A structural diagram of a computing device cluster in an embodiment of the present application;

[0046] Fig.18 This is another structural diagram of a computing device cluster in an embodiment of the present application. DETAILED DESCRIPTION

[0047] See also Figure 1 The method for obtaining encrypted short video experience index data of the present application can be applied to a wireless communication system. The wireless communication system of the present application can also be called an operator network, which includes but is not limited to access network equipment, core network equipment and customer experience management (CEM) devices. The probe belongs to the core network equipment, or the probe is a device for collecting data transmitted between core network devices. The customer experience management device can be implemented by a server or a computing device cluster.

[0048] The Internet includes over the top (OTT) servers and dial-up test servers. OTT services refer to various application services provided by third-party service providers to users through the Internet. This service is provided by third-party service providers other than operators, so third-party service providers can directly provide services and bill users. The dial-up test server is used to provide dial-up test services (i.e., dial-up test services), and various services can be tested based on the dial-up test function.

[0049] After the OTT server sends the encrypted short video code stream to the terminal, when the dial test server and the terminal perform a dial test, the terminal sends the dial test data to the dial test server, and the dial test data includes but is not limited to the short video resolution, etc. During the dial test, the OTT server sends the encrypted short video code stream to the terminal through the operator network, so that the probe of the operator network can obtain the network feature data based on the encrypted short video code stream and send the network feature data to the customer experience management device.

[0050] The following is an introduction to the customer experience management device. Figure 2 In one embodiment, the customer experience management device 200 includes a receiving module 201 and a processing module 202. The receiving module 201 is used to receive the network feature data of the business period sent by the probe, and the processing module 202 is used to input the network feature data of the business period into the artificial intelligence model, and output the resolution type of the encrypted short video in the business period through the artificial intelligence model.

[0051] In an optional embodiment, the processing module 202 is used to divide the network feature data of the service period into network feature data of multiple service sub-periods; input the network feature data of the service sub-periods into an artificial intelligence model, and output the resolution type of the encrypted short video in the service sub-period through the artificial intelligence model;

[0052] The customer experience management device 200 also includes an experience indicator module 203, which is used to determine, for each business sub-period, that the resolution type of the encrypted short video in the business sub-period is a low-definition sub-period, and determine the proportion of the target type according to the total number of low-definition sub-periods and the total number of business sub-periods; and / or, obtain the experience indicator module 203 for determining that the business sub-period is a high-definition sub-period when the resolution type of the encrypted short video in the business sub-period is a high-definition type, and determine the proportion of the high-definition type according to the total number of high-definition sub-periods and the total number of business sub-periods.

[0053] In another optional embodiment, the customer experience management device 200 also includes a policy module 204, which is used to generate a cell network construction plan according to the cell corresponding to the cell identifier when the proportion of high-definition types in the business period is less than a first threshold; when the proportion of low-definition types in the business period is greater than a second threshold, generate a cell network optimization strategy according to the cell corresponding to the cell identifier.

[0054] In another optional embodiment, the customer experience management device 200 also includes a training module 205, which is used to receive multiple encrypted short video feature samples sent by the probe and train an artificial intelligence model based on the multiple encrypted short video feature samples.

[0055] Among them, the receiving module 201, the processing module 202, the experience index module 203, the strategy module 204 and the training module 205 can all be implemented by software, or can be implemented by hardware. Taking the processing module 202 as an example, the implementation of the processing module 202 is introduced. Similarly, the implementation of the receiving module 201, the experience index module 203, the strategy module 204 and the training module 205 can refer to the implementation of the processing module 202.

[0056] As an example of a software functional unit, the processing module 202 may include code running on a computing instance. Among them, the computing instance may include at least one of a physical host (computing device), a virtual machine, and a container. Further, the above-mentioned computing instance may be one or more. For example, the processing module 202 may include code running on multiple hosts / virtual machines / containers. It should be noted that the multiple hosts / virtual machines / containers used to run the code may be distributed in the same region (region) or in different regions. Furthermore, the multiple hosts / virtual machines / containers used to run the code may be distributed in the same availability zone (AZ) or in different AZs, each AZ including a data center or multiple data centers with similar geographical locations. Among them, usually a region may include multiple AZs.

[0057] Similarly, multiple hosts / virtual machines / containers used to run the code can be distributed in the same virtual private cloud (VPC) or in multiple VPCs. Usually, a VPC is set up in a region. For cross-region communication between two VPCs in the same region and between VPCs in different regions, a communication gateway needs to be set up in each VPC to achieve interconnection between VPCs through the communication gateway.

[0058] As an example of a hardware functional unit, the processing module 002 may include at least one computing device, such as a server, etc. Alternatively, the processing module 202 may also be a device implemented using an application-specific integrated circuit (ASIC) or a programmable logic device (PLD). The PLD may be a complex programmable logical device (CPLD), a field-programmable gate array (FPGA), a generic array logic (GAL) or any combination thereof.

[0059] The multiple computing devices included in the processing module 202 can be distributed in the same region or in different regions. The multiple computing devices included in the processing module 202 can be distributed in the same AZ or in different AZs. Similarly, the multiple computing devices included in the processing module 202 can be distributed in the same VPC or in multiple VPCs. The multiple computing devices can be any combination of computing devices such as servers, ASICs, PLDs, CPLDs, FPGAs, and GALs.

[0060] It should be noted that, in other embodiments, the processing module 202 may be used to execute Figure 5 The embodiment shown or Figure 7 In any step implemented by the customer experience management device in the embodiment shown, the receiving module 201 can be used to execute Figure 5 The embodiment shown or Figure 7 In any step implemented by the customer experience management device in the embodiment shown, the experience index module 203 can be used to perform Figure 5 The embodiment shown or Figure 7 In any step implemented by the customer experience management device in the embodiment shown, the policy module 204 can be used to execute Figure 5 The embodiment shown or Figure 7 In any step implemented by the customer experience management device in the embodiment shown, the training module 205 can be used to perform Figure 5 The embodiment shown or Figure 7In the embodiment shown, any step implemented by the customer experience management device, the steps implemented by the receiving module 201, the processing module 202, the experience indicator module 203, the strategy module 204 and the training module 205 can be specified as needed, and the steps implemented by the receiving module 201, the processing module 202, the experience indicator module 203, the strategy module 204 and the training module 205 can be respectively implemented. Figure 5 The embodiment shown or Figure 7 Different steps in the illustrated embodiment implement all functions of the customer experience management device 200 .

[0061] See also Figure 3 An embodiment of the probe provided in the present application includes a receiving module 301, a processing module 302 and a sending module 303. The receiving module 301 is used to collect the encrypted short video code stream sent by the terminal during the business period; the processing module 302 is used to determine the network feature data of the business period according to the encrypted short video code stream of the business period, and the sending module 303 is used to send the network feature data of the business period to the customer experience management device.

[0062] In an optional embodiment, the processing module 302 is specifically used to remove idle periods from the business period to obtain effective data transmission duration; remove retransmitted data packets and the header of each data packet from the encrypted short video code stream of the business period to obtain valid data; determine the slice data transmission rate according to the amount of valid data and the effective data transmission duration.

[0063] In another optional embodiment, the processing module 302 is specifically used to determine short video data packets greater than a data volume threshold from the encrypted short video code stream of the business period; divide the short video data packets of the business period into multiple data slices according to a preset slice size; and determine the slice data transmission rate according to the data volume of all data slices and the duration of all data slices.

[0064] In another optional embodiment, the processing module 302 is specifically used to obtain all transactions from the encrypted short video code stream of the business period, each transaction includes a transaction request, data corresponding to the transaction request, and a confirmation message corresponding to the transaction request; the slice data transmission rate is determined according to the transaction data volume of all transactions and the transaction duration of all transactions.

[0065] In another optional embodiment, the processing module 302 is specifically used to determine short video data packets greater than a data volume threshold from the encrypted short video code stream of the business period; divide the short video data packets of the business period into multiple data slices according to a preset slice size; and determine the delay of the data slices.

[0066] In another optional embodiment, the processing module 302 is specifically used to determine short video data packets greater than a data volume threshold from the encrypted short video code stream of the business period; divide the business period into multiple time slices according to a preset slice delay; and determine the data volume of the time slice.

[0067] In another optional embodiment, the processing module 302 is specifically used to select multiple time intervals from the business period, and the time interval is smaller than the time slice; for each time interval, when the average data transmission rate of the time interval is less than the low-speed threshold, the time interval is determined to be a low-speed time interval; and the stall ratio is determined as the ratio of the number of low-speed time intervals to the total number of time intervals.

[0068] In another optional embodiment, the processing module 302 is specifically used to select multiple time intervals from the business period; for each time interval, when the average data transmission rate of the time interval is less than the low-speed threshold, the time interval is determined to be a low-speed time interval; multiple consecutive low-speed time intervals in the business period are used as low-speed periods; and the maximum continuous low-speed duration is determined as the duration of the maximum low-speed period.

[0069] In another optional embodiment, the receiving module 301 is also used to collect business data sent by the terminal during the dialing test period; the processing module 302 is also used to determine the network feature data of the dialing test period based on the encrypted short video code stream of the dialing test period; the receiving module 301 is also used to receive the dialing test data sent by the terminal; the processing module 302 is also used to match the network feature data and the resolution type according to the user identifier and the service identifier; according to the same time period in the playback period and the dialing test period, select the target network feature data and the target resolution type from the successfully matched network feature data and the resolution type; generate an encrypted short video feature sample according to the target network feature data and the target resolution type; the sending module 303 sends multiple encrypted short video feature samples to the customer experience management device.

[0070] The receiving module 301, the processing module 302 and the sending module 303 can all be implemented by software or by hardware. In other embodiments, the processing module 302 can be used to execute Figure 5 The embodiment shown or Figure 7 In any step performed by the probe in the embodiment shown, the receiving module 301 can be used to perform Figure 5 The embodiment shown or Figure 7 In any step performed by the probe in the embodiment shown, the sending module 303 can be used to perform Figure 5 The embodiment shown or Figure 7 In the embodiment shown, any step performed by the probe, the steps implemented by the receiving module 301, the processing module 302 and the sending module 303 can be specified as needed, and the steps implemented by the receiving module 301, the processing module 302 and the sending module 303 can be specified as needed. Figure 5 The embodiment shown or Figure 7 The various steps in the illustrated embodiment implement the overall functionality of the probe 300 .

[0071] Before this application uses an artificial intelligence model to process the network feature data of encrypted short videos, it is necessary to train the artificial intelligence model. The following is an introduction to the artificial intelligence model involved in this application. The basic principle of artificial intelligence (AI) is to combine massive data with super computing processing capabilities and intelligent algorithms to establish an AI model that solves specific problems, so that the AI ​​model can automatically summarize and learn potential patterns or features from the data, thereby achieving a way of thinking close to that of humans.

[0072] AI model, also known as AI algorithm (or AI operator), is a general term for mathematical algorithms built on the principles of artificial intelligence, and is also the basis for using AI to solve specific problems. Depending on the specific methods and / or technologies for implementing artificial intelligence, AI models can also be specifically referred to as machine learning models, deep learning models, or reinforcement learning models. Machine learning is a method for implementing artificial intelligence. The goal of this method is to design and analyze some algorithms (i.e., models) that allow computers to automatically "learn". The designed algorithms are called machine learning models. Machine learning models are a type of algorithm that automatically analyzes and obtains rules from data and uses rules to predict unknown data. Machine learning models include a variety of models. Depending on whether the model training needs to rely on the labels corresponding to the training data, the machine learning model involved in this application is a supervised learning model. A supervised learning model is a model obtained by determining the parameters of the initial AI model based on the data in a given training data set and the labels corresponding to each data in the training data set. The process of determining the parameters of the initial AI model using the data in the training data set and the labels corresponding to the data is also called supervised learning (or supervised training). The labels of the data in the training data set are automatically marked by probes to identify the correct answers to the data for specific tasks. Typical supervised learning models include: support vector machine, neural network model, logistic regression model, decision tree, naive Bayes model, Gaussian discriminant model, etc. Supervised learning models are usually used for classification or regression.

[0073] Deep learning is a new technical field that emerged in the process of machine learning research. Specifically, deep learning is a method in machine learning based on deep representation learning of data. Deep learning interprets data by establishing a neural network that simulates the human brain for analysis and learning. Deep learning algorithms try to learn features from data themselves. Algorithms designed based on deep learning ideas are called deep learning models.

[0074] The typical structure of deep learning models at present is a deep neural network. A neural network is a mathematical model or computational model that imitates the structure and function of a biological neural network (the central nervous system of an animal, especially the brain). A neural network is calculated by connecting a large number of neurons. A neural network can include multiple neural network layers with different functions, each of which includes parameters and calculation rules. Different layers in a neural network have different names according to different calculation formulas or functions. For example, the layer that performs convolution calculations is called a convolution layer, which is often used to extract features from input signals (such as images). A neural network can also be composed of multiple sub-neural networks. Neural networks with different structures can be applied to different scenarios (for example, classification, recognition) or provide different effects when used in the same scenario. The different structures of neural networks specifically include one or more of the following: different number of layers in the neural network, different order of each network layer, and different weights, parameters or calculation formulas in each network layer. There are many different neural networks with high accuracy for application scenarios such as recognition or classification in the industry. Some neural networks can be trained with specific data sets and used alone to complete a task or combined with other neural networks (or other functional modules) to complete a task. Classic deep learning models include convolutional neural network (CNN), recurrent neural network (RNN), recursive neural network (RNN), etc.

[0075] AI model training refers to the process of using a specified initial model to calculate the training data, and then using a certain method to adjust the parameters in the initial model based on the calculation results, so that the model gradually learns certain rules and has specific functions. After training, the AI ​​model with stable functions can be used for reasoning. AI model reasoning is the process of using the trained AI model to calculate the input data and obtain the predicted reasoning results.

[0076] In the training phase, it is first necessary to build a training set for the deep learning model based on the goal. The training set includes multiple training data, each of which is set with a label. The label of the training data is the correct answer to the specific question of the training data. The label can represent the goal of training the deep learning model using the training data. For example: to train a deep learning model that can be used to identify different animals, the training set can include multiple images of different animals (i.e., training data), and each image can have a label to identify the type of animal contained therein, such as: cat, dog. In this example, the type of animal corresponding to each image is the label of the training data.

[0077] See also Figure 4 When training a deep learning model, the training data can be input into the deep learning model after parameter initialization in batches. The deep learning model calculates (i.e., inferences) the training data to obtain prediction results for the training data. The prediction results obtained through inference and the labels corresponding to the training data are used as data for calculating the loss according to the loss function. The loss function is a function used to calculate the gap (i.e., loss value) between the prediction results of the model for the training data and the labels of the training data during the model training phase. The loss function can be implemented using different mathematical functions. Commonly used loss function expressions include: mean square error loss function, logarithmic loss function, least squares method, etc.

[0078] The loss value calculated based on the loss function can be used to update the parameters of the deep learning model. The specific parameter update method often uses the gradient descent method. Model training is a repeated iterative process. Each iteration infers different training data and calculates the loss value. The goal of multiple iterations is to continuously update the parameters of the deep learning model and find the parameter configuration that minimizes or stabilizes the loss value of the loss function.

[0079] In the training phase, in order to improve the training efficiency and performance of the model after training, it is necessary to set some reasonable hyperparameters for training. The hyperparameters of deep learning models refer to a type of parameters that cannot be obtained by learning training data during training or cannot be changed by training data. It is a concept relative to the parameters in the model. The hyperparameters of deep learning models are usually set manually based on experience or experiments. Hyperparameters include: learning rate, batch size, network structure hyperparameters (for example: number of network layers (also called depth), interaction between network layers, number and size of convolution kernels, activation function), etc. Among them, the learning rate is used as a hyperparameter to control the amplitude of the parameter weight update of the model during training, which greatly affects the speed and accuracy of training.

[0080] The trained deep learning model can be used to infer the input data. In the inference stage, the data of the actual application scenario is usually used as the input data. The reasoning of the trained deep learning model can obtain the inference result. The inference stage is the actual application of the trained deep learning model, which can quickly use the capabilities of AI to solve specific technical problems. Nowadays, AI has many application scenarios, and the reasoning of deep learning models can also be used in various application scenarios, such as: personnel recognition scenarios for access control security systems, video pornography and violence detection, express waybill number detection and recognition, etc.

[0081] The above only introduces the training of the most typical deep learning model as an example. The training of other types of models is slightly different, but the principles are similar. Most of them are to infer the training data and adjust the parameters in the model according to the inference results, with the goal of obtaining a parameter combination that stabilizes the model performance.

[0082] The following is an introduction to the process of training the artificial intelligence model in this application. Referring to FIG. 5 , an embodiment of the method for obtaining encrypted short video experience index data in this application includes:

[0083] Step 501: The probe collects service data of the dialing test period sent by the terminal.

[0084] In this embodiment, the terminal has a dial-up test function, and various services can be tested based on the dial-up test function. During the dial-up test of the encrypted short video, the terminal sends service data to the OTT server through the operator network, and the probe of the operator network can obtain the service data, which includes but is not limited to the encrypted short video stream, user identification, service time period and service identification. The encrypted short video stream of this application refers to the data stream related to the transmission of encrypted short videos between the OTT server and the terminal, which includes but is not limited to encrypted short video data, short video description information (such as short video title, short video summary, etc.).

[0085] User identification includes but is not limited to international mobile equipment identity (IMEI). Service identification includes but is not limited to short video service identification and application identification.

[0086] Step 502: The probe determines the network characteristic data of the dialing period according to the encrypted short video code stream of the dialing period. The network characteristic data includes at least two of the slice data transmission rate, the delay of the data slice, the data volume of the time slice, the maximum continuous low speed duration and the stall ratio. It should be noted that step 502 is similar to step 702, and the specific description can be referred to.

[0087] Step 503: The probe collects the dialing data sent by the terminal.

[0088] The dial test data includes multiple records, which include user ID, service period, service ID and resolution type.

[0089] Step 504: The probe matches the network feature data and the resolution type according to the user identifier and the service identifier.

[0090] Specifically, when the user identifier included in the dial test data is consistent with the user identifier included in the service data and the service identifier included in the dial test data is consistent with the service identifier included in the service data, it is determined that the resolution type included in the dial test data matches the network characteristic data corresponding to the service data successfully. When any of the following situations occurs, such as the user identifier included in the dial test data is inconsistent with the user identifier included in the service data, and the service identifier included in the dial test data is inconsistent with the service identifier included in the service data, it is determined that the resolution type included in the dial test data matches the network characteristic data corresponding to the service data unsuccessfully.

[0091] Step 505: The probe selects target network feature data and target resolution type from the successfully matched network feature data and resolution types according to the same time period in the business time period and the dialing test time period.

[0092] It should be noted that step 504 is a method for matching network feature data and resolution type with user identification and service identification, and step 505 is a method for matching network feature data and resolution type with time. Step 505 can be executed before step 504, and this application does not limit the order of the two steps.

[0093] Step 506: The probe generates an encrypted short video feature sample according to the target network feature data and the target resolution type. The resolution type in the encrypted short video feature sample is used as a sample label.

[0094] Step 507: The probe sends multiple encrypted short video feature samples to the customer experience management device.

[0095] Step 508: The customer experience management device trains an artificial intelligence model based on multiple encrypted short video feature samples.

[0096] In this embodiment, the operator's probe can match the dialing data and business data to obtain encrypted short video feature samples, and train an artificial intelligence model for identifying short video resolution based on multiple encrypted short video feature samples.

[0097] For ease of understanding, the following Figure 6 A method for associating network feature data with short video resolution is introduced. In an example, the process includes:

[0098] Step 601: The terminal dials up the short video service.

[0099] Step 602: The terminal sends a dialing test log to a dialing test server.

[0100] Step 603: The probe obtains keywords and short video resolution from the dial test log.

[0101] Keywords include user ID, service ID, and dialing time period.

[0102] Step 604: The terminal sends keywords and encrypted short video code stream to the short video server.

[0103] Step 605: The probe obtains network feature data based on the encrypted short video stream.

[0104] Step 606: The probe associates the network feature data and the short video resolution according to the keyword.

[0105] After the AI ​​model is trained, it can be used to identify the resolution type corresponding to the network feature data of the operator network. Figure 7 , an embodiment of the method for obtaining encrypted short video experience index data in the present application includes:

[0106] Step 701: The probe collects business data sent by the terminal during the business period.

[0107] Business data includes encrypted short video code stream, user ID, business ID and business time period.

[0108] Step 702: The probe determines network characteristic data of the business period according to the encrypted short video stream of the business period.

[0109] The network characteristic data includes at least two of the slice data transmission rate, the data slice delay, the data volume of the time slice, the maximum continuous low-speed duration and the stall ratio.

[0110] Step 703: The probe sends the network characteristic data of the service period to the customer experience management device.

[0111] Step 704: The customer experience management device inputs the network feature data of the business period into the artificial intelligence model, and outputs the resolution type of the encrypted short video during the business period through the artificial intelligence model.

[0112] Specifically, the resolution types include but are not limited to 480P (i.e. low-definition type), 720P (i.e. high-definition type), 1080P (i.e. full-HD type), 2K type and 4K type. The resolution type output by the model and the number of resolution types can be set according to actual conditions and are not limited in this application.

[0113] In this embodiment, the probe of the operator's network can obtain the network feature data of the encrypted short video, and the operator's customer experience management device can input the network feature data of the encrypted short video into the artificial intelligence model, and identify the resolution type of the encrypted short video through the artificial intelligence model, thus solving the problem that the operator cannot identify the resolution type of the encrypted short video.

[0114] The following is an introduction to a method for determining network characteristic data based on an encrypted short video code stream. In an optional embodiment, step 702 includes: removing idle periods from a business period to obtain effective data transmission duration, removing retransmitted data packets and the header of each data packet from the encrypted short video code stream to obtain valid data; determining a slice data transmission rate based on the amount of valid data and the effective data transmission duration.

[0115] Since the encrypted short video code stream is represented by multiple data slices during the data transmission process, and each data slice has a large gap, the average rate calculated based on the duration of multiple data slices and the service period cannot accurately reflect the data transmission rate of the data slice. In this embodiment, after removing the idle period from the service period, the effective data transmission duration of the data slice is obtained. The idle period is a period greater than the packet interval threshold and no data is transmitted. Optionally, the packet interval threshold can be but is not limited to the maximum round-trip delay between the terminal and the OTT server.

[0116] Removing retransmitted data packets can avoid the impact of retransmitted data on the slice data transmission rate and improve the accuracy of statistical valid data. The header of the data packet includes one or more of a layer 1 header, a layer 2 header, and a layer 3 header, and the specific content includes a protocol field and an IP five-tuple, etc. Since the header of the data packet does not include short video content and short video description information, removing the header of the data packet can make the amount of valid data closer to the actual amount of short video code stream data during the business period. The slice data transmission rate obtained by dividing the amount of valid data by the effective data transmission duration can accurately reflect the data transmission rate of the data slice. It should be noted that before obtaining valid data, data packets that are less than or equal to the data volume threshold can be removed.

[0117] See also Figure 8 In one example, after removing the idle period from the service period, valid data segment 1 and valid data segment 2 are obtained. The transmission duration of valid data segment 1 is t1, and the transmission duration of valid data segment 2 is t2. Then the slice data transmission rate = (the amount of data in valid data segment 1 + the amount of data in valid data segment 2) / (t1+t2).

[0118] In another optional embodiment, step 702 includes: determining short video data packets greater than a data volume threshold from the encrypted short video code stream of the business period; dividing the short video data packets of the business period into multiple data slices according to a preset slice size; and determining the slice data transmission rate according to the data volume of all data slices and the duration of all data slices.

[0119] The size of each short video data packet is greater than the data volume threshold. The data volume threshold does not exceed the upper limit of the data packet size (i.e., 1500 bytes), and its value may be, but is not limited to, 1000 bytes, and may be specifically set according to the actual size of the short video data packet. Optionally, the result obtained by dividing the data volume of all data slices by the duration of all data slices is the slice data transmission rate.

[0120] In this embodiment, short video data packets can be screened out according to the data volume threshold, and signaling data packets such as ACK can be removed, so as to obtain the accurate data volume of the short video code stream. The slice data transmission rate calculated based on the data volume of the short video code stream is more accurate than the data transmission rate calculated based on all data packets in the business period.

[0121] See also Fig. 9 In one example, the amount of data above the time axis is less than or equal to the data amount threshold, and the amount of data below the time axis is greater than the data amount threshold. Multiple data slices can be divided according to the slice size, for example, data slice 1 and data slice 2. The transmission time of data slice 1 is t1, and the transmission time of data slice 2 is t2. Then the slice data transmission rate = (the amount of data in data slice 1 + the amount of data in data slice 2) / (t1+t2).

[0122] In another optional embodiment, step 702 includes: obtaining all transactions from the encrypted short video code stream of the business period, each transaction including a transaction request, data corresponding to the transaction request, and a confirmation message corresponding to the transaction request; determining the slice data transmission rate based on the transaction data volume of all transactions and the transaction duration of all transactions.

[0123] The transaction request may include but is not limited to a hypertext transfer protocol (HTTP) request, and does not include a transmission control protocol (TCP) connection request.

[0124] Optionally, the result obtained by dividing the transaction data volume of all transactions by the transaction duration of all transactions is used as the slice data transmission rate.

[0125] See also Fig.10 In one example, the amount of data above the time axis is less than or equal to the data amount threshold, and the amount of data below the time axis is greater than the data amount threshold. The transmission duration of transaction 1 is t1, and the transmission duration of transaction 2 is t2, then the slice data transmission rate = (the amount of data of transaction 1 + the amount of data of transaction 2) / (t1+t2).

[0126] Alternatively, the data transmission rate of each transaction is calculated based on the data volume and transaction duration of the transaction, and then the average transaction rate is calculated based on the weighted sum of the data transmission rates of all transactions, and the average transaction rate is used as the slice data transmission rate.

[0127] In this embodiment, since the transaction does not include signaling for establishing a communication connection, the impact of the above signaling on the slice data transmission rate can be reduced, thereby improving the accuracy of the slice data transmission rate.

[0128] The network characteristic data may also include xKB delay, which refers to the time taken to transmit xKB. The following is an introduction to a method for obtaining xKB delay based on an encrypted short video stream in a business period. In an optional embodiment, step 702 includes: determining a short video data packet greater than a data volume threshold from an encrypted short video stream in a business period; dividing the short video data packet in the business period into a plurality of data slices according to a preset slice size; and determining the delay of the data slices.

[0129] In this application, the preset slice size can be any value in the range of 100KB to 1MB, such as 400KB, 800KB, and can be set according to actual conditions. This application can divide the short video data packets of the business period into multiple data slices according to the preset slice size. The data volume of each data slice is the preset slice size, but the duration of the data slice may be different. Fig.11 In one example, data slices can be divided according to a preset slice size and the delay of the data slices can be obtained. The delay of the data slices may include a maximum delay and a minimum delay.

[0130] The network characteristic data may also include the XMS data volume, which refers to the amount of data transmitted in the XMS. In another optional embodiment, step 702 includes: determining short video data packets greater than a data volume threshold from the encrypted short video code stream of the service period; dividing the service period into multiple time slices according to the preset slice delay; and determining the data volume of the time slice. The preset slice delay can be set according to actual conditions and is not limited in this application. Fig.12 In one example, time slices can be divided according to a preset slice delay and the data volume of the time slices can be obtained.

[0131] The xKB delay (i.e., the delay of the data slice) and the xms data volume (i.e., the data volume of the time slice) are related to the short video resolution and the short video smoothness, and therefore the download quality (i.e., the short video transmission quality) is evaluated.

[0132] In another optional embodiment, step 702 includes: selecting multiple time intervals from the business period, the time interval being smaller than the time slice; for each time interval, when the average data transmission rate of the time interval is less than the low-speed threshold, determining the time interval as a low-speed time interval; determining the stall ratio as the ratio of the number of low-speed time intervals to the total number of time intervals.

[0133] In this embodiment, the division of the time interval can be linear or nonlinear. The duration of the time interval can be set according to the actual situation, and this application does not limit it. The time interval is smaller than the time slice, which provides network feature data based on a smaller granularity related to the download quality. The duration of the time interval is any value between 50 and 300 milliseconds, for example, 100 milliseconds.

[0134] See also Fig.13 In one example, the business period is divided into time interval 1 to time interval 5, time interval 1 is a low-speed time interval, and time interval 2 to time interval 5 are not low-speed time intervals, then the stall ratio is 0.2.

[0135] It should be understood that the present application can determine a time interval as a high-speed time interval when the average data transmission rate of the time interval is greater than a high-speed threshold, and determine the high-speed proportion as the ratio of the number of high-speed time intervals to the total number of time intervals.

[0136] In another optional embodiment, step 702 includes: selecting multiple time intervals from the business period; for each time interval, when the average data transmission rate of the time interval is less than the low-speed threshold, determining the time interval as a low-speed time interval; taking multiple consecutive low-speed time intervals in the business period as low-speed periods; determining the maximum continuous low-speed duration as the duration of the maximum low-speed period. The maximum low-speed duration is related to low resolution. Fig.14 In one example, after determining the low-speed time interval in the business period, the low-speed period can be determined based on multiple consecutive low-speed time intervals, and the duration of the maximum low-speed period is used as the maximum continuous low-speed duration.

[0137] In an optional embodiment, the method of obtaining encrypted short video experience index data of the present application also includes: when the proportion of high-definition types in the business period is less than a first threshold, generating a cell network construction plan according to the cell corresponding to the cell identifier; when the proportion of low-definition types in the business period is greater than a second threshold, generating a cell network optimization strategy according to the cell corresponding to the cell identifier.

[0138] In this embodiment, when the proportion of high-definition types in the business period is less than the first threshold, it indicates that the network conditions of the cell are good for a short time and it belongs to the traffic suppression area. This application can generate a cell network construction plan, based on which the base station can be built to improve the cell traffic. When the proportion of low-definition types in the business period is greater than the second threshold, it indicates that the cell belongs to a cell with poor network quality, and the network of the cell can be optimized according to the cell network optimization strategy. The values ​​of the first threshold and the second threshold can be set according to the actual situation.

[0139] For ease of understanding, the following Fig.15 For an introduction to the method of obtaining encrypted short video experience index data in this application, see Fig.15 ,In an example, in the sample feature extraction stage, the ,automatic dialing test tool is used to obtain the dialing test log, ,read the dialing test record from the dialing test log, and ,automatically mark the feature samples of the dialing test record, ,which feature samples of the dialing test record include information such as ,short video resolution.

[0140] The terminal can also obtain the basic feature set and high-order key quality indicators of the encrypted short video code stream. The basic feature set is the feature calculated based on the data collected from the network, such as rate, data volume, delay, performance, packet interval set, etc. The high-order key quality indicators are obtained by processing the basic feature set according to the method of this application, such as slice data transmission rate, data slice delay, time slice data volume, maximum continuous low-speed duration and stall ratio, etc.

[0141] The present application can also classify the feature paths of the feature samples after marking the feature samples of the dial test record, mine features related to the short video resolution from the feature sample set, perform feature selection on the feature mining results, and calculate high-order key quality indicators based on the feature selection results. After keyword matching the high-order key quality indicators and resolution, an encrypted short video feature sample can be generated.

[0142] During the business experience fitting stage, AI model training can be performed based on multiple encrypted short video feature samples, and the trained AI model can be used to infer network feature data. The inferred experience indicator data includes resolution type, low-definition ratio, high-definition ratio, etc.

[0143] The present application also provides a computing device 1600, such as Fig.16 As shown, computing device 1600 includes: bus 1602, processor 1604, memory 1606 and communication interface 1608. Processor 1604, memory 1606 and communication interface 1608 communicate through bus 1602. Computing device 1600 can be a server or a terminal device. It should be understood that the present application does not limit the number of processors and memories in computing device 1600.

[0144] The bus 1602 may be a peripheral component interconnect (PCI) bus or an extended industry standard architecture (EISA) bus. The bus may be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Fig.16 The bus 1604 may include a path for transmitting information between various components of the computing device 1600 (eg, the memory 1606, the processor 1604, and the communication interface 1608).

[0145] The processor 1604 may include any one or more of a central processing unit (CPU), a graphics processing unit (GPU), a microprocessor (MP), or a digital signal processor (DSP).

[0146] The memory 1606 may include a volatile memory, such as a random access memory (RAM). The processor 1604 may also include a non-volatile memory, such as a read-only memory (ROM), a flash memory, a hard disk drive (HDD), or a solid state drive (SSD).

[0147] The memory 1606 stores executable program codes. In one embodiment, the processor 1604 executes the executable program codes to respectively implement Figure 2 In the embodiment shown, the processing module, the experience index module, the training module and the strategy module are implemented to realize the method of obtaining encrypted short video experience index data. In another embodiment, the processor 1604 executes the executable program code to respectively implement Figure 3 The function of the processing module in the illustrated embodiment is to implement the method for obtaining encrypted short video experience index data. That is, the memory 1606 stores instructions for executing the method for obtaining encrypted short video experience index data.

[0148] The communication interface 1608 uses a transceiver module such as, but not limited to, a network interface card or a transceiver to implement communication between the computing device 1600 and other devices or communication networks. In one embodiment, the communication interface 1608 can implement Figure 2 In another embodiment, the communication interface 1608 can implement Figure 3 The functions of the receiving module and the sending module in the illustrated embodiment.

[0149] The embodiment of the present application also provides a computing device cluster. The computing device cluster includes at least one computing device. The computing device can be a server, such as a central server, an edge server, or a local server in a local data center. In some embodiments, the computing device can also be a terminal device such as a probe, a desktop, a laptop, or a smart phone.

[0150] like Fig.17 As shown, the computing device cluster includes at least one computing device 1600. The memory 1606 in one or more computing devices 1600 in the computing device cluster may store the same instructions for executing the method for obtaining encrypted short video experience index data.

[0151] In some possible implementations, the memory 1606 of one or more computing devices 1600 in the computing device cluster may also respectively store partial instructions for executing the method for obtaining encrypted short video experience index data. In other words, the combination of one or more computing devices 1600 may jointly execute instructions for executing the method for obtaining encrypted short video experience index data.

[0152] It should be noted that the memory 1606 in different computing devices 1600 in the computing device cluster can store different instructions, which are respectively used to execute part of the functions of the customer experience management device. That is, the instructions stored in the memory 1606 in different computing devices 1600 can implement the functions of one or more modules of the processing module, the experience indicator module, the training module and the strategy module.

[0153] In some possible implementations, one or more computing devices in the computing device cluster may be connected via a network, which may be a wide area network or a local area network. Fig.18 A possible implementation is shown. Fig.18As shown, two computing devices 1600A and 1600B are connected via a network. Specifically, they are connected to the network via a communication interface in each computing device. In this type of possible implementation, the memory 1606 in the computing device 1600A stores instructions for executing the functions of the processing module, the experience index module, and the strategy module. At the same time, the memory 1606 in the computing device 1600B stores instructions for executing the functions of the training module.

[0154] The embodiment of the present application also provides a computer program product including instructions. The computer program product may be a software or program product including instructions that can be run on a computing device or stored in any available medium. When the computer program product is run on at least one computing device, the at least one computing device executes a method for obtaining encrypted short video experience indicator data.

[0155] The embodiment of the present application also provides a computer-readable storage medium. The computer-readable storage medium can be any available medium that can be stored by a computer or a data storage device such as a data center containing one or more available media. The available medium can be a magnetic medium (e.g., a floppy disk, a hard disk, a tape), an optical medium (e.g., a DVD), or a semiconductor medium (e.g., a solid-state hard disk). The computer-readable storage medium includes instructions that instruct the computer to execute a method for obtaining encrypted short video experience indicator data.

[0156] The present application also provides a chip system, which includes a processor and a memory coupled to each other. The memory is used to store computer programs or instructions, and the processor is used to execute the computer programs or instructions stored in the memory, so that the computing device executes the steps performed by the processing module, experience index module, training module and strategy module in the above embodiments. Optionally, the memory is a memory in the chip, such as a register, a cache, etc., and the memory can also be a memory located outside the chip in the site, such as a read-only memory or other types of static storage devices that can store static information and instructions, a random access memory, etc. The processor mentioned in any of the above places can be a general-purpose central processing unit, a microprocessor, a dedicated integrated circuit or one or more integrated circuits for implementing the above-mentioned method for obtaining encrypted short video experience index data.

[0157] The above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit it. Although the present application has been described in detail with reference to the aforementioned embodiments, a person skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present application.

Claims

1. A method for obtaining encrypted short video experience index data, characterized in that: include: Receive network feature data of a service period sent by a probe, wherein the network feature data of the service period is obtained by the probe according to an encrypted short video code stream of the service period, and the network feature data includes at least two items of a slice data transmission rate, a data slice delay, a time slice data volume, a maximum continuous low-speed duration, and a stall ratio; The network feature data of the business period is input into an artificial intelligence model, and the resolution type of the encrypted short video during the business period is output through the artificial intelligence model.

2. The method according to claim 1, characterized in that The step of inputting the network feature data of the service period into an artificial intelligence model and outputting the resolution type of the encrypted short video during the service period through the artificial intelligence model comprises: Dividing the network characteristic data of the service period into network characteristic data of a plurality of service sub-periods; Inputting the network feature data of the service sub-period into an artificial intelligence model, and outputting the resolution type of the encrypted short video in the service sub-period through the artificial intelligence model; The method further comprises: For each of the service sub-periods, when the resolution type of the encrypted short video in the service sub-period is a low-definition type, determining the service sub-period as a low-definition sub-period; determining the proportion of the low-definition type according to the total number of the low-definition sub-periods and the total number of the service sub-periods; and / or, For each of the business sub-periods, when the resolution type of the encrypted short video in the business sub-period is a high-definition type, the business sub-period is determined to be a high-definition sub-period; and the proportion of high-definition types is determined based on the total number of high-definition sub-periods and the total number of business sub-periods.

3. The method according to claim 2, characterized in that The network characteristic data also includes a cell identifier, and the method further includes: When the proportion of high-definition types in the service period is less than a first threshold, generating a cell network construction plan according to the cell corresponding to the cell identifier; When the proportion of low-definition types in the service period is greater than the second threshold, a cell network optimization strategy is generated according to the cell corresponding to the cell identifier.

4. The method according to any one of claims 1 to 3, characterized in that Before receiving the network characteristic data of the service period sent by the probe, the method further includes: Receive multiple encrypted short video feature samples sent by the probe, wherein the encrypted short video feature samples include network feature data of the encrypted short video and a resolution type of the encrypted short video, and the resolution type of the encrypted short video is a sample label; The artificial intelligence model is trained based on a plurality of the encrypted short video feature samples.

5. A method for obtaining short video experience index data, characterized in that: include: Collect the encrypted short video stream sent by the terminal during business hours; Determine network characteristic data of the service period according to the encrypted short video code stream of the service period, wherein the network characteristic data includes at least two items of slice data transmission rate, data slice delay, time slice data volume, maximum continuous low speed duration and stall ratio; The network characteristic data of the service period is sent to a customer experience management device.

6. The method according to claim 5, characterized in that Determining the network characteristic data according to the encrypted short video code stream of the business period includes: Removing an idle period from the service period to obtain an effective data transmission duration, wherein the idle period is a period greater than a packet interval threshold and during which no data is transmitted; Remove the retransmitted data packets and the header of each data packet from the encrypted short video code stream of the service period to obtain valid data; The slice data transmission rate is determined according to the data volume of the valid data and the valid data transmission duration.

7. The method according to claim 5, characterized in that Determining the network characteristic data according to the encrypted short video code stream of the business period includes: Determine a short video data packet having a data volume greater than a data volume threshold from the encrypted short video code stream in the business period; Dividing the short video data packets of the service period into a plurality of data slices according to a preset slice size; The slice data transmission rate is determined according to the data volume of all data slices and the duration of all data slices.

8. The method according to claim 5, characterized in that Determining the network characteristic data according to the encrypted short video code stream of the business period includes: Acquire all transactions from the encrypted short video code stream of the business period, each transaction including a transaction request, data corresponding to the transaction request, and a confirmation message corresponding to the transaction request; The slice data transmission rate is determined based on the transaction data volume of all transactions and the transaction duration of all transactions.

9. The method according to claim 5, characterized in that Determining the network characteristic data according to the encrypted short video code stream of the business period includes: Determine a short video data packet having a data volume greater than a threshold value from the encrypted short video code stream of the service period; Dividing the short video data packets of the service period into a plurality of data slices according to a preset slice size; A latency of the data slice is determined.

10. The method according to claim 5, characterized in that Determining the network characteristic data according to the encrypted short video code stream of the business period includes: Determine a short video data packet having a data volume greater than a data volume threshold from the encrypted short video code stream in the business period; Dividing the service period into multiple time slices according to a preset slice delay; The data amount of the time slice is determined.

11. The method according to claim 5, characterized in that Determining the network characteristic data according to the encrypted short video code stream of the business period includes: Selecting a plurality of time intervals from the business period, wherein the time interval is smaller than the time slice; For each time interval, when the average data transmission rate of the time interval is less than a low-speed threshold, the time interval is determined to be a low-speed time interval; The stall ratio is determined as the ratio of the number of low-speed time intervals to the total number of time intervals.

12. The method according to claim 5, characterized in that Determining the network characteristic data according to the encrypted short video code stream of the business period includes: Selecting multiple time intervals from the business period; For each time interval, when the average data transmission rate of the time interval is less than a low-speed threshold, the time interval is determined to be a low-speed time interval; taking a plurality of consecutive low-speed time intervals in the business period as a low-speed period; The maximum continuous low speed duration is determined as the duration of the maximum low speed period.

13. The method according to any one of claims 5 to 12, characterized in that Before the encrypted short video code stream is sent by the acquisition terminal, the method further includes: Collecting service data sent by the terminal during the dialing test period, the service data including the encrypted short video code stream, user identifier, service period and service identifier; Determine network characteristic data of the dial test period according to the encrypted short video code stream of the dial test period; Receiving dialing test data sent by a terminal, the dialing test data including a plurality of records, the records including a user identifier, a dialing test period, a service identifier and a resolution type; Matching the network characteristic data with the resolution type according to the user identifier and the service identifier; According to the same period in the business period and the dialing test period, selecting target network feature data and target resolution type from the successfully matched network feature data and resolution types; Generate an encrypted short video feature sample according to the target network feature data and the target resolution type; Send the plurality of encrypted short video feature samples to the customer experience management device.

14. A customer experience management device, characterized in that: include: A receiving module, used to receive network characteristic data of a business period sent by a probe, wherein the network characteristic data of the business period is obtained by the probe according to the encrypted short video code stream of the business period, and the network characteristic data includes at least two items of a slice data transmission rate, a data slice delay, a time slice data volume, a maximum continuous low-speed duration, and a stall ratio; A processing module is used to input the network feature data of the business period into an artificial intelligence model, and output the resolution type of the encrypted short video during the business period through the artificial intelligence model.

15. The device according to claim 14, characterized in that The processing module is specifically used to divide the network feature data of the service period into network feature data of multiple service sub-periods; input the network feature data of the service sub-periods into an artificial intelligence model, and output the resolution type of the encrypted short video in the service sub-period through the artificial intelligence model; The device also includes an experience indicator module, which is used to determine, for each of the service sub-periods, when the resolution type of the encrypted short video in the service sub-period is a low-definition type, that the service sub-period is a low-definition sub-period, and determine the proportion of the low-definition type according to the total number of the low-definition sub-periods and the total number of the service sub-periods; And / or, for each of the business sub-periods, when the resolution type of the encrypted short video in the business sub-period is a high-definition type, the business sub-period is determined to be a high-definition sub-period, and the proportion of the high-definition type is determined according to the total number of the high-definition sub-periods and the total number of the business sub-periods.

16. The device according to claim 14, characterized in that The network feature data also includes a cell identifier, and the device also includes a strategy module, which is used to generate a cell network construction plan according to the cell corresponding to the cell identifier when the proportion of high-definition types in the business period is less than a first threshold; when the proportion of low-definition types in the business period is greater than a second threshold, generate a cell network optimization strategy according to the cell corresponding to the cell identifier.

17. The device according to any one of claims 14 to 16, characterized in that The device also includes a training module, which is used to receive multiple encrypted short video feature samples sent by the probe, the encrypted short video feature samples include network feature data of the encrypted short video and the resolution type of the encrypted short video, and the resolution type of the encrypted short video is a sample label; the artificial intelligence model is trained according to the multiple encrypted short video feature samples.

18. A probe, characterized in that: include: A receiving module, used to collect the encrypted short video stream sent by the terminal during the business period; A processing module, configured to determine network characteristic data of a service period according to the encrypted short video code stream of the service period, wherein the network characteristic data includes at least two items of a slice data transmission rate, a data slice delay, a time slice data volume, a maximum continuous low-speed duration, and a stall ratio; The sending module is used to send the network characteristic data of the service period to the customer experience management device.

19. The probe according to claim 18, characterized in that The processing module is specifically used to remove idle periods from the business period to obtain effective data transmission duration, wherein the idle period is a period greater than a packet interval threshold and during which no data is transmitted; remove retransmitted data packets and the header of each data packet from the encrypted short video code stream of the business period to obtain effective data; and determine the slice data transmission rate according to the amount of effective data and the effective data transmission duration.

20. The probe according to claim 18, characterized in that The processing module is specifically used to determine a short video data packet greater than a data volume threshold from the encrypted short video code stream of the business period; divide the short video data packet of the business period into multiple data slices according to a preset slice size; The slice data transmission rate is determined according to the data volume of all data slices and the duration of all data slices.

21. The probe according to claim 18, characterized in that The processing module is specifically used to obtain all transactions from the encrypted short video code stream of the business period, each of which includes a transaction request, data corresponding to the transaction request, and a confirmation message corresponding to the transaction request; The slice data transmission rate is determined based on the transaction data volume of all transactions and the transaction duration of all transactions.

22. The probe according to claim 18, characterized in that The processing module is specifically used to determine short video data packets greater than a data volume threshold from the encrypted short video code stream of the business period; divide the short video data packets of the business period into multiple data slices according to a preset slice size; and determine the delay of the data slices.

23. The probe according to claim 18, characterized in that The processing module is specifically used to determine short video data packets greater than a data volume threshold from the encrypted short video code stream of a business period; divide the business period into multiple time slices according to a preset slice delay; and determine the data volume of the time slice.

24. The probe according to claim 18, characterized in that The processing module is specifically used to select multiple time intervals from the business period, and the time interval is smaller than the time slice; for each time interval, when the average data transmission rate of the time interval is less than the low-speed threshold, the time interval is determined to be a low-speed time interval; and the stall ratio is determined as the ratio of the number of low-speed time intervals to the total number of time intervals.

25. The probe according to claim 18, characterized in that The processing module is specifically used to select multiple time intervals from the business period; for each time interval, when the average data transmission rate of the time interval is less than the low-speed threshold, determine that the time interval is a low-speed time interval; taking a plurality of consecutive low-speed time intervals in the business period as a low-speed period; The maximum continuous low speed duration is determined as the duration of the maximum low speed period.

26. The probe according to any one of claims 18 to 25, characterized in that The receiving module is also used to collect business data sent by the terminal during the dialing test period, and the business data includes an encrypted short video stream, a user identifier, a business period and a business identifier; The processing module is also used to determine the network feature data of the dialing test period according to the encrypted short video code stream of the dialing test period; The receiving module is also used to receive the dialing test data sent by the terminal, the dialing test data includes multiple records, and the records include a user identifier, a dialing test period, a service identifier and a resolution type; The processing module is also used to match the network feature data and the resolution type according to the user identifier and the service identifier; select the target network feature data and the target resolution type from the successfully matched network feature data and the resolution type according to the same period in the service period and the dialing period; Generate an encrypted short video feature sample according to the target network feature data and the target resolution type; and send a plurality of the encrypted short video feature samples to a customer experience management device.

27. A computing device, characterized in that The method comprises a processor and a memory, wherein the processor is configured to execute instructions stored in the memory so that the computing device executes the method according to any one of claims 1 to 13.

28. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a computing device, the computing device performs the method according to any one of claims 1 to 13.

29. A computer program product comprising instructions, characterized in that When the instructions are executed by a computing device, the computing device is caused to perform the method according to any one of claims 1 to 13.

Images