Data transmission method and device, computer equipment and readable storage medium

Abstract

The invention relates to a data transmission method and device, computer equipment, a computer readable storage medium and a computer program product. The method comprises the following steps: acquiring a network attribute data set and a historical user behavior data set of each user group, and acquiring a to-be-recommended video data set of each user group; determining a predicted sliding time sequence of each user group according to the prediction model, each network attribute data set and each historical user behavior data set; the predicted sliding time sequence represents the probability that the user group slides the to-be-recommended video in the to-be-recommended video data set; taking the minimum total service delay of each user group as a target, and constructing a service delay model according to the predicted sliding time sequence, the to-be-recommended video data set and the network attribute data set; and determining a target transmission bandwidth based on the service delay model, and transmitting the to-be-recommended video data set based on the target transmission bandwidth. By adopting the method, the playing fluency of the video data can be improved.

Description

Technical Field

[0001] The present application relates to the field of video data technology, and in particular to a data transmission method, apparatus, computer equipment, computer-readable storage medium, and computer program product. Background Art

[0002] In real life, more and more people watch short videos in their spare time. In order to ensure that users can watch short videos, the edge server needs to transmit the video data of the short video to the user terminal through a data transmission method so that the user terminal can load the video data of the short video.

[0003] In the current data transmission method, when the video data cached in the terminal is played, the user terminal sends a video data request in response to the user terminal to the edge server. The edge server determines the target transmission bandwidth based on the current network attribute data, and sends the data set to be transmitted to the user's terminal based on the target transmission bandwidth.

[0004] However, in the current data transmission method, users will slide the video data to speed up the video data playback, which results in the transmission speed of the recommended video not keeping up with the playback speed of the cached video data. Therefore, the current data transmission method has a delay and the video data playback is not smooth. Summary of the invention

[0005] Based on this, it is necessary to provide a data transmission method, apparatus, computer device, computer-readable storage medium and computer program product to address the above technical issues.

[0006] In a first aspect, the present application provides a data transmission method, comprising:

[0007] Obtaining a network attribute dataset and a historical user behavior dataset for each user group, and obtaining a dataset of videos to be recommended for each user group;

[0008] Determine a predicted sliding time series for each user group according to the prediction model, each network attribute data set and each historical user behavior data set; the predicted sliding time series represents the probability of a user group sliding a video to be recommended in the video data set to be recommended;

[0009] Taking the total service delay of each user group as the minimum as the goal, a service delay model is constructed according to the predicted sliding time series, the to-be-recommended video dataset and the network attribute dataset;

[0010] A target transmission bandwidth is determined based on the service delay model, and the to-be-recommended video data set is transmitted based on the target transmission bandwidth.

[0011] In one embodiment, obtaining a network attribute data set and a historical user behavior data set for each user group includes:

[0012] For each user group, according to a preset historical time period, obtain the historical sliding time series, geographical area and interest tags of each user in the user group;

[0013] Constructing a historical user behavior dataset of the user group according to the historical sliding time series of each of the users, the geographical area, and the interest tags;

[0014] Acquire edge server attribute data, user terminal attribute data of each of the users, and a base station attribute data set;

[0015] A network attribute data set is constructed based on the edge server attribute data, the attribute data of each user terminal and the attribute data sets of each base station.

[0016] In one embodiment, determining the predicted sliding time series of each user group according to the prediction model, the network attribute data set and the historical user behavior data set includes:

[0017] Based on the prediction model, prediction processing is performed on the network attribute features in the network attribute data set and the historical user behavior features in the historical user behavior data set of each user group to obtain a predicted sliding time series of the user group.

[0018] In one embodiment, the network attribute data set includes edge server attribute data, and the service delay model is constructed based on the predicted sliding time series, the video data set to be recommended, and the network attribute data set with the goal of minimizing the total service delay of each user group, including:

[0019] For each of the user groups, constructing a video transcoding delay function according to the to-be-recommended video dataset, the predicted sliding time series, a weight coefficient and edge server attribute data;

[0020] Constructing a multicast transmission delay function according to the Shannon algorithm, the predicted sliding time series and the edge server attribute data;

[0021] With the goal of minimizing the total service delay of each user group, a service delay model is constructed according to bandwidth resource data, the video transcoding delay function and the multicast transmission delay function.

[0022] In one embodiment, for each user group, constructing a video transcoding delay function according to the to-be-recommended video dataset, the predicted sliding time series, a weight coefficient and edge server attribute data includes:

[0023] For each of the user groups, integrating the to-be-recommended video data set based on the predicted sliding time series to obtain average viewing video data of the user group;

[0024] A video transcoding delay function is constructed according to the average viewing video data, the weight coefficient, and the computing capability data of the edge server in the edge server attribute data.

[0025] In one embodiment, the minimizing the total service delay of each user group is taken as the goal, and the service delay model is constructed according to the bandwidth resource data, the video transcoding delay function and the multicast transmission delay function, including:

[0026] Establishing constraint conditions based on bandwidth resource data of each of the user groups;

[0027] Taking the minimum total service delay of each user group as the goal, constructing a total service delay function according to the video transcoding delay function and the multicast transmission delay function;

[0028] The total service delay function and the constraint condition are combined to obtain the service delay model.

[0029] In a second aspect, the present application also provides a data transmission device, including:

[0030] An acquisition module is used to acquire a network attribute data set and a historical user behavior data set of each user group, and acquire a video data set to be recommended of each user group;

[0031] A determination module, used to determine a predicted sliding time series of each user group according to a prediction model, each network attribute data set and each historical user behavior data set; the predicted sliding time series represents the probability of a user group sliding a video to be recommended in the video data set to be recommended;

[0032] A construction module, configured to construct a service delay model according to the predicted sliding time series, the to-be-recommended video dataset and the network attribute dataset, with the goal of minimizing the total service delay of each user group;

[0033] A transmission module is used to determine a target transmission bandwidth based on the service delay model, and transmit the to-be-recommended video data set based on the target transmission bandwidth.

[0034] In a third aspect, the present application further provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the following steps are implemented:

[0035] Obtaining a network attribute dataset and a historical user behavior dataset for each user group, and obtaining a dataset of videos to be recommended for each user group;

[0036] Determine a predicted sliding time series for each user group according to the prediction model, each network attribute data set and each historical user behavior data set; the predicted sliding time series represents the probability of a user group sliding a video to be recommended in the video data set to be recommended;

[0037] Taking the total service delay of each user group as the minimum as the goal, a service delay model is constructed according to the predicted sliding time series, the to-be-recommended video dataset and the network attribute dataset;

[0038] A target transmission bandwidth is determined based on the service delay model, and the to-be-recommended video data set is transmitted based on the target transmission bandwidth.

[0039] In a fourth aspect, the present application further provides a computer-readable storage medium having a computer program stored thereon, wherein when the computer program is executed by a processor, the following steps are implemented:

[0040] Obtaining a network attribute dataset and a historical user behavior dataset for each user group, and obtaining a dataset of videos to be recommended for each user group;

[0041] Determine a predicted sliding time series for each user group according to the prediction model, each network attribute data set and each historical user behavior data set; the predicted sliding time series represents the probability of a user group sliding a video to be recommended in the video data set to be recommended;

[0042] Taking the total service delay of each user group as the minimum as the goal, a service delay model is constructed according to the predicted sliding time series, the to-be-recommended video dataset and the network attribute dataset;

[0043] A target transmission bandwidth is determined based on the service delay model, and the to-be-recommended video data set is transmitted based on the target transmission bandwidth.

[0044] In a fifth aspect, the present application further provides a computer program product, including a computer program, which implements the following steps when executed by a processor:

[0045] Obtaining a network attribute dataset and a historical user behavior dataset for each user group, and obtaining a dataset of videos to be recommended for each user group;

[0046] Determine a predicted sliding time series for each user group according to the prediction model, each network attribute data set and each historical user behavior data set; the predicted sliding time series represents the probability of a user group sliding a video to be recommended in the video data set to be recommended;

[0047] Taking the total service delay of each user group as the minimum as the goal, a service delay model is constructed according to the predicted sliding time series, the to-be-recommended video dataset and the network attribute dataset;

[0048] A target transmission bandwidth is determined based on the service delay model, and the to-be-recommended video data set is transmitted based on the target transmission bandwidth.

[0049] The above-mentioned data transmission method, device, computer equipment, computer-readable storage medium and computer program product obtain network attribute data sets and historical user behavior data sets of each user group, and obtain the video data sets to be recommended of each user group; determine the predicted sliding time series of each user group according to the prediction model, the network attribute data set and the historical user behavior data set; the predicted sliding time series characterizes the probability of the user group sliding the video to be recommended in the video data set to be recommended; with the goal of minimizing the total service delay of each user group, a service delay model is constructed according to the predicted sliding time series, the video data set to be recommended and the network attribute data set; determine the target transmission bandwidth based on the service delay model, and transmit the video data set to be recommended based on the target transmission bandwidth. With this method, the predicted sliding time series is determined by the prediction model, the network attribute data set and the historical user behavior data set, the probability of the user group sliding the video to be recommended is clarified, and the target transmission bandwidth is determined based on the predicted sliding time series, the video data set to be recommended and the network attribute data set, and the probability of the user group sliding the video to be recommended is added. The factor can be used to more reasonably allocate the target transmission bandwidth for transmitting the video to be recommended, and the accuracy of the target transmission bandwidth is improved. Furthermore, the video data set to be recommended is transmitted based on the target transmission bandwidth, which improves the transmission speed of the video data set to be recommended, reduces data transmission delay, and improves the fluency of playing video data. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] In order to more clearly illustrate the technical solutions in the embodiments of the present application or related technologies, the drawings required for use in the embodiments of the present application or related technical descriptions will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without paying creative work.

[0051] Figure 1 is an application environment diagram of a data transmission method in an embodiment;

[0052] Figure 2 A schematic diagram of a flow chart of a data transmission method in an embodiment;

[0053] Figure 3A schematic diagram of a process for obtaining each user group and each historical user behavior data set in one embodiment;

[0054] Figure 4 A schematic diagram of a process for constructing a service delay model in an embodiment;

[0055] Figure 5 A schematic diagram of a process for constructing a video transcoding delay function in one embodiment;

[0056] Figure 6 A schematic diagram of a process for constructing a service delay model in an embodiment;

[0057] Figure 7 A schematic diagram of a flow chart of a data transmission method in another embodiment;

[0058] Figure 8 is a system architecture diagram of a data transmission device in an exemplary embodiment;

[0059] Fig. 9 is a structural block diagram of a data transmission device in one embodiment;

[0060] Fig.10 FIG. 4 is a diagram showing the internal structure of a computer device in one embodiment. DETAILED DESCRIPTION

[0061] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0062] The data transmission method provided in the embodiment of the present application can be applied to Figure 1The data transmission system 100 shown in the figure. The data transmission system 100 includes a user terminal 110, a base station 120 and an edge server 130. The edge server 130 stores various video data and sends the various video data to the user terminal 110 of the user through the base station 120. The user terminal 110 and the base station 120 are connected through a network, and the edge server 130 and the base station 120 are also connected through a network. The data storage system can store the data that the edge server 130 needs to process. The data storage system can be integrated on the edge server 130, or it can be placed on the cloud or other network servers. Among them, the user terminal 110 can be, but is not limited to, various personal computers, laptops, smart phones, tablet computers, Internet of Things devices and portable wearable devices. The Internet of Things devices can be smart speakers, smart TVs, smart air conditioners, smart car devices, projection devices, etc. Portable wearable devices can be smart watches, smart bracelets, head-mounted devices, etc. The head-mounted devices can be virtual reality (VR) devices, augmented reality (AR) devices, smart glasses, etc. The edge server 130 may be an independent physical server, or a server cluster or distributed system composed of multiple physical servers, or a cloud server providing cloud computing services.

[0063] In an exemplary embodiment, Figure 2 As shown, a data transmission method is provided, which is applied to Figure 1 The edge server 130 (hereinafter, the reference numeral is omitted, referred to as the edge server) in the embodiment is used as an example to illustrate the method, which includes the following steps 202 to 208. Among them:

[0064] Step 202: Obtain a network attribute data set and a historical user behavior data set of each user group, and obtain a video data set to be recommended of each user group.

[0065] The users in each user group have the same video browsing characteristics.

[0066] In implementation, the edge server obtains network attribute data of each user in each user group to obtain a network attribute data set. The edge server obtains historical user behavior data of each user in the user group according to a preset historical time period to obtain a historical behavior data set. Then, the edge server obtains a video data set to be recommended for the user group.

[0067] Specifically, the edge server obtains the user portrait of each user. Then, the edge server divides the users with the same user portrait into a user group to obtain each user group. The edge server obtains the network attribute data of the network where each user in the user group is located from the database to obtain a network attribute data set. At the same time, the edge server obtains the historical user behavior data of each user in the user group from the database according to a preset historical time period to obtain a historical behavior data set. Then, the edge server selects the video data set to be recommended corresponding to the user group from each video data in the database according to the user portrait of the user group. The video data set to be recommended is the video data sent to each user in the user group.

[0068] In an optional embodiment, a video recommendation model is pre-set in the edge server. The edge server obtains a historical browsing video data set browsed by each user in the user group within a historical time period. The edge server inputs the historical browsing video data set and each video data into the video recommendation model, and screens each video data to be recommended from each video data through the video recommendation model to obtain the video data set to be recommended.

[0069] Optionally, the user grouping basis may be, but is not limited to, user portraits or user interest tags, and the screening basis of the video data set to be recommended may be user portraits or user interest tags, both of which represent the user's video browsing characteristics. The present application embodiment does not limit the user grouping basis and the screening basis.

[0070] Optionally, the user group may be any user group among the user groups, and the embodiment of the present application does not limit the user group.

[0071] Step 204 , determining a predicted sliding time series for each user group based on the prediction model, the network attribute data set, and the historical user behavior data set.

[0072] The predicted sliding time series represents the probability of a user group sliding a video to be recommended in the video dataset to be recommended.

[0073] In implementation, the edge server obtains the network attribute data set and the historical user behavior data set for each user group and inputs them into the prediction model. The prediction model performs prediction processing on the network attribute data set and the historical user behavior data set to obtain the predicted sliding time series of the user group. By determining the predicted sliding time series, the probability of each user in the user group being in the sliding video dataset to be recommended is obtained.

[0074] Step 206 , with the goal of minimizing the total service delay of each user group, a service delay model is constructed according to the predicted sliding time series, the video dataset to be recommended, and the network attribute dataset.

[0075] In the implementation, the edge server constructs a video transcoding delay function and a multicast transmission delay function based on the predicted sliding time series, the video dataset to be recommended, and the network attribute dataset. Then, the edge server takes the minimum total service delay of each user group as the goal, and constructs a service delay model based on the video transcoding delay function, the multicast transmission delay function, and the network attribute dataset. The service delay model characterizes the delay when the edge server sends the video dataset to be recommended to the user terminal of the user group.

[0076] Step 208: determine a target transmission bandwidth based on the service delay model, and transmit the video data set to be recommended based on the target transmission bandwidth.

[0077] In implementation, the edge server solves the service delay model according to a preset solver to obtain the target transmission bandwidth of each user group. Then, the edge server transmits the video data set to be recommended according to the target transmission bandwidth of each user group.

[0078] Specifically, a linear solver is pre-set in the edge server, and the edge server solves the service delay model through the linear solver to obtain the target transmission bandwidth of each user group. Then, for each user group, the edge server sends the recommended video data set to the base station corresponding to each user in the user group according to the target transmission bandwidth of the user group, and sends the video data set to be recommended to the user terminal of each user through the base station. The user terminal of each user receives the video data set to be recommended and caches it locally.

[0079] In the above data transmission method, the predicted sliding time series is determined through the prediction model, the network attribute data set and the historical user behavior data set, the probability of the user group sliding the video to be recommended is clarified, and the target transmission bandwidth is determined based on the predicted sliding time series, the video data set to be recommended and the network attribute data set, adding the factor of the probability of the user group sliding the video to be recommended, which can more reasonably allocate the target transmission bandwidth for transmitting the video to be recommended and improve the accuracy of the target transmission bandwidth. Furthermore, the video data set to be recommended is transmitted based on the target transmission bandwidth, which improves the speed of transmitting the video data set to be recommended, reduces the data transmission delay, and improves the fluency of playing video data.

[0080] In an exemplary embodiment, Figure 3 As shown, the specific processing process of obtaining the network attribute data set and the historical user behavior data set of each user group in step 202 includes steps 302 to 308. Among them:

[0081] Step 302, for each user group, according to a preset historical time period, obtain the historical sliding time series, geographical area and interest tags of each user in the user group.

[0082] The historical time period includes each historical time slice. The historical sliding time series represents the probability of a user sliding video data in each historical time slice.

[0083] In implementation, the edge server is pre-set with a historical time period. The historical time period corresponds to the predicted time period. The edge server obtains the user's historical sliding time series, geographic area, and interest tags from the database for each user in each user group.

[0084] In an exemplary embodiment, the historical time period is the 30 minutes before the current time. The predicted time period is the next 10 minutes after the current time. For each user in each user group, the edge server obtains the user's historical sliding time series, geographic area, and interest tags from the database for the 30 minutes before the current time according to the user ID of the user. The geographic area represents the user's current location, and the interest tag represents the user's video browsing characteristics.

[0085] Optionally, the historical time period may be, but is not limited to, the 30 minutes before the current time. The historical time period is determined based on forecast requirements, and the embodiments of the present application do not limit the historical time period.

[0086] Step 304 , construct a historical user behavior dataset of the user group based on the historical sliding time series, geographic area, and interest tags of each user.

[0087] In implementation, the edge server combines the historical sliding time series, geographic region and behavior tag of each user to obtain the historical user behavior data of the user. Then, the edge server constructs the historical behavior data set of the user group based on the historical behavior data of each user.

[0088] Optionally, the user's historical user behavior data set includes at least a historical sliding time series, a geographic area, and an interest tag. The embodiment of the present application does not limit the historical user behavior data in the historical user behavior data set.

[0089] Step 306: Acquire edge server attribute data, user terminal attribute data of each user, and base station attribute data set.

[0090] In implementation, the edge server obtains the edge server attribute data of the current edge server. Then, the edge server obtains the user terminal attribute data and the base station attribute data of each user.

[0091] Specifically, the edge server obtains the bandwidth, computing power, cache, and computing workload of the edge server, and combines the bandwidth, computing power, cache, and computing workload of the edge server to obtain edge server attribute data of the edge server. Then, for each user in the user group, the edge server obtains the network quality data of the base station where the user is located and the network transmission performance data of the user terminal of the user. The edge server determines the network quality data of the base station where the user is located as the base station attribute data, and determines the network transmission performance data of the user terminal as the user terminal attribute data.

[0092] Optionally, the user terminal attribute data includes at least network transmission performance data of the user terminal, and the base station attribute data includes at least network quality data of the base station where the user is located. The user terminal attribute data and the base station attribute data are determined based on predicted demand. The embodiment of the present application does not limit the user terminal attribute data and the base station attribute data.

[0093] Step 308: construct a network attribute data set based on the edge server attribute data, the attribute data of each user terminal and the attribute data set of each base station.

[0094] In implementation, the edge server combines the edge server attribute data, the user terminal attribute data and the base station attribute data set into a network attribute data subset of the user for each user in the user group. Then, the edge server combines the network attribute data subsets of each user to obtain the network attribute data set of the user group.

[0095] In this embodiment, by acquiring the network attribute data set and historical user behavior data set of the user group, the network attribute characteristics and historical user behavior characteristics of the user are obtained, which facilitates the subsequent determination of the predicted sliding time series of the user group. In addition, by determining the prediction model, simulating the real network status and user behavior dynamics of the user group, predicting the user behavior in real time and calculating the error between the expected target and the actual result, the user group model is dynamically updated, the sliding probability distribution and the recommended video list are output, the video content is prepared and transmitted in advance, and the user waiting time is reduced.

[0096] In an exemplary embodiment, the specific processing process of step 204 includes:

[0097] Based on the prediction model, the network attribute features in the network attribute data set of each user group and the historical user behavior features in the historical user behavior data set are predicted and processed to obtain the predicted sliding time series of the user group.

[0098] Among them, the network attribute characteristics and historical user behavior characteristics of the user group are correlated with the probability of the user group sliding the video data to be recommended.

[0099] In implementation, the prediction model can learn the correlation between network attribute features and historical user behavior features and the probability of a user group sliding the video data to be recommended. For each user group, the edge server inputs the network attribute data set and the historical user behavior data set of the user group into the prediction model, extracts features from the network attribute data set through the prediction model to obtain network attribute features, and extracts features from the historical behavior data set to obtain historical user behavior features. Then, the prediction model in the edge server determines the predicted sliding time series of the user group within the prediction time period based on the correlation between the network attribute features and the historical user behavior features and the probability of the user group sliding the video data to be recommended. The prediction time period contains each prediction time slice. The predicted sliding time series contains the probability of the user group sliding the video to be recommended in the video data set to be recommended in each prediction time slice.

[0100] In an exemplary embodiment, the network attribute data set characterizes the situation of the edge server transmitting video data to the user terminal. The network situation represented by the network attribute data set and the user's historical behavior state represented by the user's historical behavior data set will affect the predicted sliding time series of the user group. The historical time period is the first 30 minutes starting from the current time. The predicted time period is the last 10 minutes starting from the current time. The edge server inputs the network attribute data set of each user group and the historical user behavior data set of the first 30 minutes starting from the current time into the prediction model, and predicts the network attribute data set and the historical behavior data set through the prediction model to obtain the predicted sliding time series of the user group in the last 10 minutes starting from the current time.

[0101] In an optional embodiment, an initial prediction model is pre-set in the edge server, and the initial prediction model is an RNN (Recurrent Neural Network) algorithm. The edge server obtains a sample data set. The sample data set includes each sample data subset. Each sample data subset includes a sample network attribute data subset, a sample historical user behavior data set, and a sample sliding time series of a sample user group. For each sample data subset, the edge server inputs the sample network attribute data subset and the sample historical user behavior data subset in the sample data subset into the initial prediction model, and performs prediction processing on the sample network attribute data subset and the sample historical user behavior data subset through the initial prediction model to obtain a sample prediction result. The edge server performs data processing on the sample prediction result and the sample sliding time series according to a preset loss function to obtain a loss value. Then, the edge server determines whether the loss value reaches a preset loss threshold. If the loss value does not reach the loss threshold, the edge server updates the parameters in the initial prediction model according to the loss value, and executes the step of inputting the sample network attribute data subset and the sample historical user behavior data subset in the sample data subset into the initial prediction model until the loss value reaches the loss threshold. If the loss value reaches the loss threshold, the edge server determines the initial prediction model as the prediction model.

[0102] Optionally, the training stop condition of the initial prediction model may be but is not limited to the loss value reaching a loss threshold. The loss threshold is determined according to training requirements. The embodiment of the present application does not limit the training stop condition and loss threshold of the initial prediction model.

[0103] Optionally, the initial prediction model may be, but is not limited to, an RNN algorithm, and the embodiments of the present application do not limit the initial prediction model.

[0104] In this embodiment, the prediction model is used to determine the predicted sliding time series of the user group based on the network attribute data and the historical user behavior data set, and the probability of the user group sliding the video to be recommended is clarified, which facilitates the subsequent determination of the target bandwidth of the user group.

[0105] In an exemplary embodiment, Figure 4 As shown, the specific processing process of step 206 includes steps 402 to 406. Among them:

[0106] Step 402: for each user group, a video transcoding delay function is constructed according to the video data set to be recommended, the predicted sliding time series, the weight coefficient and the edge server attribute data.

[0107] In implementation, the edge server needs to transcode the video data to be recommended before sending the video data set to be recommended, so as to send the transcoded video data set to be recommended. For each user group, the edge server determines the average video data watched by the user group based on the video data set to be recommended and the predicted sliding time series. The average video data watched represents the average video size of the video data set to be recommended that the user group watches. Then, the edge server constructs a video transcoding delay function based on the average video data watched, the weight coefficient and the edge server attribute data. The video transcoding delay function represents the time required for transcoding the video data set to be recommended.

[0108] Step 404: construct a multicast transmission delay function according to the Shannon algorithm, the predicted sliding time series and the edge server attribute data.

[0109] Among them, the multicast transmission delay function represents the multicast transmission delay. The multicast transmission delay refers to the time required for the recommended video data set to be transmitted from the edge server to the user terminal of each user, and is also the time required to transmit the entire recommended video data set, that is, the video size divided by the transmission rate.

[0110] In the implementation, the edge server is pre-set with the Shannon algorithm. The multicast transmission delay is related to the multicast transmission rate rg(Bg) and the average transmitted video size. Among them, rg is the symbol transmission rate, Symbol Rate. The edge server processes the predicted sliding time series and edge server attribute data according to the Shannon algorithm (Shannong theorem) to obtain the multicast transmission delay function.

[0111] Specifically, the edge server attribute data set includes bandwidth, channel gain and noise power. The Shannon algorithm is shown in the following formula (1):

[0112] (1)

[0113] In the above formula (1), represents the multicast transmission rate, where is the bandwidth allocated to user group g; is the minimum value of the channel gain of all users in the user group; P is the noise power.

[0114] The edge server processes the predicted sliding time series and edge server attribute data according to the Shannon algorithm (Shannon theorem) to obtain the multicast transmission delay function. The multicast transmission delay function is shown in the following formula (2):

[0115] (2)

[0116] in, Represents the video dataset to be recommended. Indicates the length of the video dataset to be recommended. Represents the predicted sliding time series of the video dataset to be recommended. Indicates the multicast transmission rate, Delay for multicast transmission.

[0117] Step 406 , with the goal of minimizing the total service delay of each user group, a service delay model is constructed according to the bandwidth resource data, the video transcoding delay function and the multicast transmission delay function.

[0118] In the implementation, the edge server constructs a total service delay function for each user group based on the video transcoding function and the multicast transmission delay function. Then, the edge server constructs constraints based on bandwidth resource data and constructs a service delay model based on the total service delay function and the constraints.

[0119] In this embodiment, a service delay model is constructed based on the predicted sliding time series, the video data set to be recommended and the network attribute data set, so that the target transmission bandwidth can be determined according to the service delay model, and the probability of the user group sliding the video to be recommended is added, so that the target transmission bandwidth for transmitting the video to be recommended can be allocated more reasonably, thereby improving the accuracy of the target transmission bandwidth.

[0120] In an exemplary embodiment, Figure 5 As shown, the specific processing process of step 402 includes steps 502 to 504. Among them:

[0121] Step 502 : for each user group, the to-be-recommended video data set is integrated based on the predicted sliding time series to obtain average viewing video data of the user group.

[0122] In implementation, the edge server integrates the to-be-recommended video data set and the predicted sliding time series of each user group according to the average viewed video data algorithm to obtain the average viewed video data of the user group.

[0123] Specifically, in order to calculate the transcoding delay of the video data set to be recommended, the average transcoding workload needs to be calculated. When calculating the average transcoding workload, it is necessary to determine the video portion of the video data set to be recommended that is actually watched by the user group. The edge server integrates the video data set to be recommended and the predicted sliding time series according to the average viewing video data algorithm to obtain the average viewing video data of the user group. The average viewing video data algorithm is shown in the following formula (3):

[0124] (3)

[0125] In the above formula (3), Represents the average video viewing data. Represents the video dataset to be recommended. Indicates the length of the video dataset to be recommended. represents the predicted sliding time series of the video dataset to be recommended. L is the number of layers of video coding, is the size of the video dataset to be recommended at layer 1.

[0126] The average video viewing data algorithm means that the edge server multiplies the viewing probability of each part in the video data set to be recommended by the size of the part according to the predicted sliding time series, and integrates the entire video data to be recommended, that is, the transcoding workload is the accumulation of all video transcoding operations in discrete time, and the average video viewing data of the user group can be obtained. The average video viewing data represents the average size of videos watched by the user group.

[0127] Step 504: construct a video transcoding delay function according to the average viewing video data, the weight coefficient, and the computing capability data of the edge server in the edge server attribute data.

[0128] In the implementation, the edge server processes the average viewing video data, the weight coefficient, and the computing power data of the edge server in the edge server attribute data according to the video transcoding delay algorithm to obtain a video transcoding delay function. The video transcoding delay algorithm is shown in the following formula (4):

[0129] (4)

[0130] In the above formula (4), is the average video viewing data, is the computing power of the edge server, is the weight coefficient, Delay for video transcoding.

[0131] In this embodiment, the video data set to be recommended is integrated by predicting the sliding time series, and a video transcoding delay function is constructed according to the average viewing video data obtained by the integration process, so that the edge server does not need to transcode all the video data sets to be recommended, which reduces the time required for transcoding and improves the accuracy of video transcoding delay.

[0132] In an exemplary embodiment, Figure 6 As shown, the specific processing process of step 406 includes steps 602 to 606. Among them:

[0133] Step 602: construct constraint conditions based on bandwidth resource data of each user group.

[0134] The constraint conditions include a first constraint condition and a second constraint condition.

[0135] In implementation, the edge server sets the goal that the sum of bandwidth resources cannot exceed the total bandwidth capacity of the edge server network, and constructs a first constraint condition based on the bandwidth resource data of each user group. Then, the edge server sets the goal that the bandwidth resources of the user group should not be negative, and constructs a first constraint condition based on the bandwidth resource data of each user group. The first constraint condition is shown in the following formula (5):

[0136] (5)

[0137] In the above formula (5), is the bandwidth resource data of each user group, Represents the total bandwidth capacity of the edge server network. represents a user group. G represents the collection of all user groups.

[0138] The second constraint is shown in formula (6):

[0139] (6)

[0140] In the above formula (6), Indicates for all user groups (including user groups). Bandwidth resources for the user group.

[0141] Step 604 , with the goal of minimizing the total service delay of each user group, a total service delay function is constructed according to the video transcoding delay function and the multicast transmission delay function.

[0142] In the implementation, the edge server performs an addition operation on the video transcoding delay function and the multicast transmission delay function to obtain an initial total service delay function. The initial total service delay function is shown in the following formula (7):

[0143] (7)

[0144] In the above formula (7), Delay function for video transcoding. Represents the multicast transmission delay function. is the initial total service delay function.

[0145] Then, the edge server updates the initial total service delay function with the goal of minimizing the total service delay of each user group to obtain the total service delay function. The total service delay function is shown in the following formula (8):

[0146] (8)

[0147] In the above formula (8), is the minimum, represents the multicast transmission delay of each g-th user group. is the total service delay function. represents a user group. G represents the collection of all user groups.

[0148] Step 606: Combine the total service delay function and the constraints to obtain a service delay model.

[0149] In implementation, the edge server will combine the total service delay function and the constraints to obtain a service delay model. The service delay model is shown in the following formula group (9):

[0150]

[0151] (9)

[0152]

[0153] Among them, in the above formula group (9), Bandwidth resource data for user groups, is the minimum value of the channel gain of all users in the user group; P is the noise power. Indicates the length of the video dataset to be recommended. represents the predicted sliding time series of the video dataset to be recommended. G represents the collection of each user group. Represents the total bandwidth capacity of the edge server network.

[0154] In this embodiment, by taking the minimum total service delay of each user group as the goal, constructing a service delay model, and determining the target transmission bandwidth based on the service delay model, the probability of the user group sliding the video to be recommended is increased, and the target transmission bandwidth for transmitting the video to be recommended can be allocated more reasonably, and the accuracy of the target transmission bandwidth is improved. Furthermore, the video data set to be recommended is transmitted based on the target transmission bandwidth, which improves the speed of transmitting the video data set to be recommended, reduces the data transmission delay, and improves the fluency of playing video data.

[0155] In an exemplary embodiment, Figure 7 FIG. 1 is a flow chart of a data transmission method in another embodiment. Figure 7 As shown, a data transmission method is provided, comprising:

[0156] Step 701: Obtain a network attribute data set and a historical user behavior data set for each user group.

[0157] Step 702 : Perform prediction processing on the network attribute data set and the historical user behavior data set of each user group based on the prediction model to generate a prediction sliding time series.

[0158] Step 703: Obtain a dataset of videos to be recommended for each user group.

[0159] Step 704 , with the goal of minimizing the total service delay of each user group, a service delay model is constructed according to the predicted sliding time series, the video dataset to be recommended, and the network attribute dataset.

[0160] Step 705: determine a target transmission bandwidth based on the service delay model, and transmit the video data set to be recommended based on the target transmission bandwidth.

[0161] Step 706: Obtain the sliding time series corresponding to the data set to be transmitted.

[0162] Step 707: Update the prediction model according to the sliding time series and the predicted sliding time series.

[0163] In an exemplary embodiment, Figure 8 FIG. 1 is a system architecture diagram of a data transmission device in an exemplary embodiment. Figure 8 As shown, the data transmission device includes a data module and a network management module. The data module includes a prediction model, a service delay model and a processing model. The network management module includes a network controller.

[0164] Among them, the prediction model is used for user simulation and analysis of user status. The data set of the prediction model comes from the real user status data of the physical mobile communication network, which is divided into two categories: network attribute data set (network data) and historical user behavior data set (user behavior data). Specifically, network attribute data includes the network quality of the base station where the user is located, the cache and computing workload of the edge server, the network transmission performance of the mobile device, etc. The historical user behavior data set includes historical sliding time series, geographic areas, interest tags (content preferences), etc. The prediction model uses a time series-based deep learning algorithm, such as the RNN algorithm, inputs the network attribute data set and the historical user behavior data set, and generates predictions for future time steps.

[0165] Service delay model. It takes into account video transcoding delay and multicast transmission delay, monitors the delay in the video transcoding process in real time, and analyzes factors that may affect the smoothness of video playback. Among them, video transcoding delay refers to the amount of video data that can be processed per unit time. The higher the transcoding rate, the shorter the time required to complete the same transcoding amount. The transcoding amount is calculated as the integral of the probability of a user watching each part of the video and the size of the video in a discrete time period; multicast transmission delay refers to the time required for video data to be transmitted from the server to multiple viewers, which can be defined as the time required to transmit the entire video collection, that is, the video size divided by the transmission rate. It is related to the multicast transmission rate and the average transmitted video size.

[0166] The data processing model is based on the prediction model and service delay model. It can calculate the transcoding rate in real time to adapt to the network environment of different users and achieve high-throughput multicast transmission. It analyzes the demand and network conditions of the video stream and dynamically adjusts the video transcoding rate to avoid buffer vacancy, thereby improving the smoothness of the video stream. It analyzes user behavior patterns and predicts popular video segments, which can more effectively allocate transcoding resources and reduce unnecessary transcoding work.

[0167] The network controller adjusts network bandwidth allocation, routing selection and other parameters in real time based on data feedback from the data module.

[0168] The data transmission device takes into account the impact of video transcoding and multicast transmission on service delay, and implements an optimization algorithm focused on the allocation of bandwidth resources to reduce the video streaming service delay.

[0169] It should be understood that, although the steps in the flowcharts involved in the above embodiments are displayed in sequence according to the indication of the arrows, these steps are not necessarily executed in sequence according to the order indicated by the arrows. Unless there is a clear explanation in this article, the execution of these steps is not strictly limited in order, and these steps can be executed in other orders. Moreover, at least a part of the steps in the flowcharts involved in the above embodiments may include multiple steps or multiple stages, and these steps or stages are not necessarily executed at the same time, but can be executed at different times, and the execution order of these steps or stages is not necessarily carried out in sequence, but can be executed in turn or alternately with other steps or at least a part of the steps or stages in other steps.

[0170] Based on the same inventive concept, the embodiment of the present application also provides a data transmission device for implementing the data transmission method involved above. The implementation scheme for solving the problem provided by the device is similar to the implementation scheme recorded in the above method, so the specific limitations in one or more data transmission device embodiments provided below can refer to the limitations on the data transmission method above, and will not be repeated here.

[0171] In an exemplary embodiment, Fig. 9 As shown, a data transmission device 900 is provided, comprising: an acquisition module 901, a determination module 902, a construction module 903 and a transmission module 904, wherein:

[0172] The acquisition module 901 is used to acquire a network attribute data set and a historical user behavior data set of each user group, and acquire a video data set to be recommended of each user group;

[0173] The determination module 902 is used to determine the predicted sliding time series of each user group according to the prediction model, each network attribute data set and each historical user behavior data set; the predicted sliding time series represents the probability of the user group sliding the video to be recommended in the video data set to be recommended;

[0174] A construction module 903 is used to construct a service delay model according to the predicted sliding time series, the video data set to be recommended and the network attribute data set with the goal of minimizing the total service delay of each user group;

[0175] The transmission module 904 is configured to determine a target transmission bandwidth based on the service delay model, and transmit the to-be-recommended video data set based on the target transmission bandwidth.

[0176] In an exemplary embodiment, the acquisition module 901 includes a first acquisition submodule and a second acquisition submodule. The first acquisition submodule includes:

[0177] The third acquisition submodule is used to acquire, for each user group, the historical sliding time series, geographical area and interest tags of each user in the user group according to a preset historical time period;

[0178] The first construction submodule is used to construct a historical user behavior dataset of a user group based on the historical sliding time series, geographic area and interest tags of each user;

[0179] A fourth acquisition submodule, used to acquire edge server attribute data, user terminal attribute data of each user, and a base station attribute data set;

[0180] The second construction submodule is used to construct a network attribute data set based on the edge server attribute data, the attribute data of each user terminal and the attribute data set of each base station.

[0181] In an exemplary embodiment, the determination module 902 is specifically used to perform prediction processing on the network attribute features in the network attribute data set and the historical user behavior features in the historical user behavior data set of each user group based on the prediction model to obtain the predicted sliding time series of the user group.

[0182] In an exemplary embodiment, the network attribute data set includes edge server attribute data, and the construction module 903 includes:

[0183] The third construction submodule is used to construct a video transcoding delay function for each user group according to the video dataset to be recommended, the predicted sliding time series, the weight coefficient and the edge server attribute data;

[0184] a fourth construction submodule, for constructing a multicast transmission delay function according to the Shannon algorithm, the predicted sliding time series and the edge server attribute data;

[0185] The fifth construction submodule is used to construct a service delay model according to bandwidth resource data, video transcoding delay function and multicast transmission delay function with the goal of minimizing the total service delay of each user group.

[0186] In an exemplary embodiment, the third building block includes:

[0187] The first processing submodule is used to perform integration processing on the recommended video data set based on the predicted sliding time series for each user group to obtain the average viewing video data of the user group;

[0188] The sixth construction submodule is used to construct a video transcoding delay function according to the average viewing video data, the weight coefficient, and the computing capacity data of the edge server in the edge server attribute data.

[0189] In an exemplary embodiment, the fifth building block comprises:

[0190] A seventh construction submodule, used to construct constraint conditions based on bandwidth resource data of each user group;

[0191] An eighth construction submodule, configured to construct a total service delay function according to the video transcoding delay function and the multicast transmission delay function with the goal of minimizing the total service delay of each user group;

[0192] The first combining submodule is used to combine the total service delay function and the constraint conditions to obtain a service delay model.

[0193] Each module in the above data transmission device can be implemented in whole or in part by software, hardware, or a combination thereof. Each module can be embedded in or independent of a processor in a computer device in the form of hardware, or can be stored in a memory in a computer device in the form of software, so that the processor can call and execute operations corresponding to each module.

[0194] In an exemplary embodiment, a computer device is provided. The computer device may be a server, and its internal structure diagram may be as shown in FIG. Fig.10As shown. The computer device includes a processor, a memory, an input / output interface (Input / Output, referred to as I / O) and a communication interface. Among them, the processor, the memory and the input / output interface are connected through a system bus, and the communication interface is connected to the system bus through the input / output interface. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store data used for the data transmission method. The input / output interface of the computer device is used to exchange information between the processor and an external device. The communication interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, a data transmission method is implemented.

[0195] Those skilled in the art will understand that Fig.10 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0196] In one embodiment, a computer device is further provided, including a memory and a processor, wherein a computer program is stored in the memory, and the processor implements the steps in the above method embodiments when executing the computer program.

[0197] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the steps in the above-mentioned method embodiments are implemented.

[0198] In one embodiment, a computer program product is provided, including a computer program, which implements the steps in the above method embodiments when executed by a processor.

[0199] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with relevant regulations.

[0200] A person of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment method can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to the memory, database or other medium used in the embodiments provided in the present application can include at least one of non-volatile memory and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. As an illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM). The database involved in each embodiment provided in this application may include at least one of a relational database and a non-relational database. Non-relational databases may include distributed databases based on blockchains, etc., but are not limited to this. The processor involved in each embodiment provided in this application may be a general-purpose processor, a central processing unit, a graphics processor, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, an artificial intelligence (AI) processor, etc., but are not limited to this.

[0201] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this application.

[0202] The above-described embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the present application. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the attached claims.

Claims

1. A data transmission method, characterized in that: The method comprises: Obtaining a network attribute dataset and a historical user behavior dataset for each user group, and obtaining a dataset of videos to be recommended for each user group; Determine a predicted sliding time series for each user group according to the prediction model, each network attribute data set and each historical user behavior data set; the predicted sliding time series represents the probability of a user group sliding a video to be recommended in the video data set to be recommended; Taking the total service delay of each user group as the minimum as the goal, a service delay model is constructed according to the predicted sliding time series, the to-be-recommended video dataset and the network attribute dataset; A target transmission bandwidth is determined based on the service delay model, and the to-be-recommended video data set is transmitted based on the target transmission bandwidth.

2. The method according to claim 1, characterized in that The step of obtaining a network attribute data set and a historical user behavior data set for each user group includes: For each user group, according to a preset historical time period, obtain the historical sliding time series, geographical area and interest tags of each user in the user group; Constructing a historical user behavior dataset of the user group according to the historical sliding time series of each of the users, the geographical area, and the interest tags; Acquire edge server attribute data, user terminal attribute data of each of the users, and a base station attribute data set; A network attribute data set is constructed based on the edge server attribute data, the attribute data of each user terminal and the attribute data sets of each base station.

3. The method according to claim 1, characterized in that The step of determining the predicted sliding time series of each user group according to the prediction model, the network attribute data set, and the historical user behavior data set includes: Based on the prediction model, prediction processing is performed on the network attribute features in the network attribute data set and the historical user behavior features in the historical user behavior data set of each user group to obtain a predicted sliding time series of the user group.

4. The method according to claim 1, characterized in that: The network attribute data set includes edge server attribute data, and the service delay model is constructed based on the predicted sliding time series, the video data set to be recommended, and the network attribute data set with the goal of minimizing the total service delay of each user group, including: For each of the user groups, constructing a video transcoding delay function according to the to-be-recommended video dataset, the predicted sliding time series, a weight coefficient and edge server attribute data; Constructing a multicast transmission delay function according to the Shannon algorithm, the predicted sliding time series and the edge server attribute data; With the goal of minimizing the total service delay of each user group, a service delay model is constructed according to bandwidth resource data, the video transcoding delay function and the multicast transmission delay function.

5. The method according to claim 4, characterized in that The step of constructing a video transcoding delay function for each user group according to the to-be-recommended video data set, the predicted sliding time series, a weight coefficient and edge server attribute data includes: For each of the user groups, integrating the to-be-recommended video data set based on the predicted sliding time series to obtain average viewing video data of the user group; A video transcoding delay function is constructed according to the average viewing video data, the weight coefficient, and the computing capability data of the edge server in the edge server attribute data.

6. The method according to claim 4, characterized in that The method of minimizing the total service delay of each user group and constructing a service delay model according to bandwidth resource data, the video transcoding delay function and the multicast transmission delay function includes: Establishing constraint conditions based on bandwidth resource data of each of the user groups; Taking the minimum total service delay of each user group as the goal, constructing a total service delay function according to the video transcoding delay function and the multicast transmission delay function; The total service delay function and the constraint condition are combined to obtain the service delay model.

7. A data transmission device, characterized in that: The device comprises: An acquisition module is used to acquire a network attribute data set and a historical user behavior data set of each user group, and acquire a video data set to be recommended of each user group; A determination module, used to determine a predicted sliding time series of each user group according to a prediction model, each network attribute data set and each historical user behavior data set; the predicted sliding time series represents the probability of a user group sliding a video to be recommended in the video data set to be recommended; A construction module, configured to construct a service delay model according to the predicted sliding time series, the to-be-recommended video dataset and the network attribute dataset, with the goal of minimizing the total service delay of each user group; A transmission module is used to determine a target transmission bandwidth based on the service delay model, and transmit the to-be-recommended video data set based on the target transmission bandwidth.

8. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 6 are implemented.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

10. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

Images