Content data acquisition method and device based on double links

Abstract

The invention discloses a content data acquisition method and device based on double links, which are used for improving the content data acquisition efficiency of communication users. The scheme provided by the invention is applied to an edge computing device MEC, and comprises the following steps: receiving a content data request of a user terminal, the content data request carrying request information of target content data requested to be acquired by the user terminal; establishing a first service link between the MEC and the cloud center based on the content data request, and instructing the user terminal to establish a second service link between the user terminal and the cloud center based on the content data request; predicting the transmission quality of the target content data by the first service link and the second service link to determine a target service link with high transmission quality in the first service link and the second service link; and controlling the user terminal to acquire at least part of the target content data through the target service link through the local user plane function UPF equipment.

Description

Technical Field

[0001] This application relates to the field of communications, and in particular to a method and apparatus for acquiring content data based on dual links. Background Technology

[0002] In the field of communication technology, Mobile Edge Computing (MEC) can efficiently meet the communication needs of users and improve system capacity. However, in practical application scenarios, edge computing has limited storage space and computing resources, and some user-requested content data is not cached in the edge server's cache. In this case, user devices often need to establish a communication connection with the cloud center to obtain the required content data. However, this method has low data transmission efficiency and a poor user experience.

[0003] How to improve the efficiency of content data acquisition for communication users is the technical problem that this application aims to solve. Summary of the Invention

[0004] The purpose of this application is to provide a content data acquisition method and apparatus based on dual links, so as to improve the efficiency of content data acquisition for communication users.

[0005] Firstly, a dual-link-based content data acquisition method is provided, applied to edge computing devices (MECs), including: Receive a content data request from a user terminal, wherein the content data request carries indication information of the target content data requested by the user terminal; Based on the content data request, a first service link is established between the MEC and the cloud center, and the user terminal is instructed to establish a second service link between the user terminal and the cloud center based on the content data request. Predict the transmission quality of the first service link and the second service link to the target content data, so as to determine the target service link with high transmission quality among the first service link and the second service link; The user terminal is controlled by the local user plane function UPF device to obtain at least part of the target content data through the target service link.

[0006] Secondly, a dual-link content data acquisition device is provided, applied to edge computing devices (MECs), including: The receiving module receives a content data request from a user terminal, wherein the content data request carries indication information of the target content data requested by the user terminal; The module establishes a first service link between the MEC and the cloud center based on the content data request, and instructs the user terminal to establish a second service link between the user terminal and the cloud center based on the content data request. The prediction module predicts the transmission quality of the target content data to the first service link and the second service link, so as to determine the target service link with high transmission quality among the first service link and the second service link. The control module controls the user terminal to obtain at least a portion of the target content data through the target service link via the local user plane function (UPF) device.

[0007] Thirdly, an electronic device is provided, comprising a processor, a memory, and a computer program stored in the memory and executable on the processor, wherein the computer program, when executed by the processor, implements the steps of the method of the first aspect.

[0008] Fourthly, a computer-readable storage medium is provided on which a computer program is stored, which, when executed by a processor, implements the steps of the method of the first aspect.

[0009] Fifthly, a computer program product is provided, comprising a non-transitory computer-readable storage medium storing a computer program operable to cause a computer to perform some or all of the steps of the method of the first aspect.

[0010] In this embodiment, the edge computing device (MEC) first receives a content data request from a user terminal, which carries indication information of the target content data requested by the user terminal. Then, a first service link is established between the MEC and the cloud center based on the content data request, and the user terminal is instructed to establish a second service link between the user terminal and the cloud center based on the content data request. Next, the transmission quality of the target content data via the first and second service links is predicted to determine the target service link with higher transmission quality. Subsequently, the user terminal is controlled to obtain at least a portion of the target content data through the target service link via the local user plane function (UPF) device. The solution provided by this embodiment enables the establishment of a first service link between the MEC and the cloud center based on the user terminal's content data request, and a second service link between the user terminal and the cloud center. Furthermore, the transmission quality of these two service links is predicted, and the user terminal is controlled to obtain the target content data from the service link with higher transmission quality via the UPF, fully utilizing link transmission resources and effectively improving the efficiency of content data acquisition for communication users. Attached Figure Description

[0011] The accompanying drawings, which are included to provide a further understanding of this application and form part of this application, illustrate exemplary embodiments and are used to explain this application, but do not constitute an undue limitation of this application. In the drawings: Figure 1a This is one of the flowcharts illustrating a dual-link content data acquisition method according to an embodiment of this application; Figure 1b This is a schematic diagram of the process of establishing a business link for a content data acquisition method based on dual links, as described in one embodiment of this application. Figure 2a This is a second schematic flowchart of an embodiment of the content data acquisition method based on dual links in this application; Figure 2b This is a third flowchart illustrating an embodiment of the content data acquisition method based on dual links in this application; Figure 3 This is a schematic flowchart of an embodiment of the present application of a content data acquisition method based on dual links; Figure 4 This is the fifth flowchart of an embodiment of the present application of a content data acquisition method based on dual links; Figure 5a This is a schematic flowchart of an embodiment of the content data acquisition method based on dual links, which is shown in Figure 6 of this application. Figure 5b This is a matrix operation diagram of a content data acquisition method based on dual links, as an embodiment of this application; Figure 5c This is a schematic diagram of the parameter feature detection process of a content data acquisition method based on dual links, as an embodiment of this application; Figure 5d This is the seventh flowchart of an embodiment of the content data acquisition method based on dual links in this application; Figure 6a This is the eighth flowchart of an embodiment of the content data acquisition method based on dual links in this application; Figure 6b This is a flowchart of an embodiment of the content data acquisition method based on dual links, shown in Figure 9 of this application. Figure 7 This is a schematic diagram of the structure of a content data acquisition device based on dual links, which is an embodiment of this application. Detailed Implementation

[0012] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application. The drawing numbers in this application are only used to distinguish the various steps in the solution and are not used to limit the execution order of the various steps. The specific execution order is subject to the description in the specification.

[0013] In the field of communication technology, user equipment can request target content data from MEC (Multi-access Edge Computing). If the target content data storage is a cached resource of MEC, MEC can directly provide the target content data to the user equipment to meet the user equipment's content data needs.

[0014] In practical applications, edge computing servers have limited storage and computing resources, making it impossible to cache all user business requests. When a user initiates a request for content resources not available on the edge computing server, the lack of a proper solution often leads to latency or resource waste for the user's business.

[0015] To address the problems existing in related technologies, embodiments of this application provide a content data acquisition method based on dual links, such as... Figure 1a As shown, the device applied to the edge computing (MEC) includes: S11: Receive a content data request from a user terminal, wherein the content data request carries indication information of the target content data requested by the user terminal.

[0016] In this step, the content data request can be sent by the user terminal to the MEC when it needs to obtain target content data. This content data request carries indication information of the target content data, which indicates the target content data that the user device is requesting to obtain, such as the address, identifier, type, etc. of the target content data.

[0017] S12: Based on the content data request, establish a first service link between the MEC and the cloud center, and instruct the user terminal to establish a second service link between the user terminal and the cloud center based on the content data request.

[0018] In this step, a first service link is established between the MEC and the cloud center. This first service link allows the MEC to offload resources to the cloud center, thereby providing the target content data obtained from the offload to the user device. Additionally, the MEC instructs the user terminal to establish a second service link to request the target content data from the cloud center through this second service link.

[0019] See Figure 1b The interaction flow shown involves the user terminal first sending a content data request to the EMC. The EMC processes the received content data request using methods such as local DNS (Domain Name System) resolution to obtain the instruction information carried within. Based on this instruction information, a first service link and a second service link are established, as shown in the dashed box. It should be noted that after the EMC receives the content data request and parses out the instruction information, it can establish the first and second service links simultaneously, or it can establish them sequentially. Figure 1b This is only used to illustrate the interaction method of the business links, and does not limit the order in which the two business links are established.

[0020] The second service link can be constructed based on the MEC's ​​UPF (User Plane Function) traffic offloading rules. For example, the local MEC AF (Application Function) pre-informs the PCF (Point Coordination Function) of the UPF traffic offloading rules via the N5 / N33 interface. The PC (Point Coordinator) then configures the traffic offloading strategy to the SMF (Session Management Function). The SMF centrally schedules all traffic and can use solutions such as LADN (Local Area Data Network), UL-CL (Uplink Classifier), or Multi-Homing traffic offloading to select traffic offloading for the edge UPF. Local traffic that needs to be offloaded is offloaded through the local edge UPF, while non-local traffic is sent to the central UPF for processing through the local UPF. This avoids all traffic detouring through the central network, reducing the pressure on the backbone network transmission and network construction costs, and improving the carrying efficiency of packet data within the network and the user service experience.

[0021] For example, see Figure 1b The steps within the dashed box indicating the establishment of the second service link are as follows: Based on the parsed instruction information, the MEC sends cloud center resource address information to the user terminal to instruct the user terminal to establish a second service link with the cloud center. Specifically, the user terminal can send a cloud center content data request to its local UPF based on the cloud center resource address information. The local UPF routes the content data request to the central UPF, which then routes it to the cloud center. Thus, upon receiving the content data request, the cloud center establishes a terminal service link (the second service link) with the user terminal.

[0022] See Figure 1b The steps within the dashed box indicating the establishment of the first service link are as follows: Based on the parsed instruction information, the MEC routes the content data request from the aforementioned user terminal to the local UPF, which then routes the content data request to the central UPF, and finally the central UPF routes the content data request to the cloud center. Thus, upon receiving the content data request, the cloud center establishes an MEC service link with the MEC, which is the first service link.

[0023] The first business link described above can be used by the MEC to unload target content data from the cloud center, and then provide the target content data to the user terminal. The second business link described above can be used for the user terminal to directly connect with the cloud center to obtain target content data directly from the cloud center. Through the solution provided in this application embodiment, when a user triggers a content data request, the request is first routed to the MEC side. The MEC can then unload resources to the cloud center before sending the request to the user, or the user can directly establish a link with the cloud center to obtain the content data. In the solution provided in this application embodiment, a dual-link establishment mechanism is used to build business links between the user, MEC, and cloud center, and both constructed links can achieve content data acquisition.

[0024] S13: Predict the transmission quality of the target content data to the first service link and the second service link to determine the target service link with high transmission quality among the first service link and the second service link.

[0025] In this step, transmission quality prediction is performed based on the established first and second service links. Specifically, the transmission parameters of these two service links can be obtained separately to predict the transmission quality of the target content data. Transmission quality includes, for example, transmission continuity, transmission efficiency, and link stability.

[0026] Specifically, the transmission quality of a service link can be predicted by summarizing collected service link parameters based on preset rules using statistical parameters. Alternatively, a large model can be applied to predict the transmission quality of a service link.

[0027] S14: Control the user terminal to obtain at least part of the target content data through the target service link via the local user plane function UPF device.

[0028] In this step, after identifying the target service link with high transmission quality, the user terminal is controlled by the local UPF device to obtain the target content data through the target service link with high quality, so as to improve the transmission efficiency of the target content data.

[0029] In this step, the local UPF device can be used to control the user terminal to initiate the acquisition of target content data through the target service link. Alternatively, it can be used to control the user terminal to switch from the current service link transmitting the target content data to a higher quality target service link.

[0030] For example, in the solution provided in this application embodiment, a default service link can be pre-defined between the first service link and the second service link. For instance, after establishing the second service link, the user terminal immediately transmits the target content data through the second service link. After determining the target service link through this solution, if the target service link is not the aforementioned default service link, the user terminal is controlled by the local UPF device to switch from the default service link to the target service link. For example, switching from the second service link that is currently transmitting the target content data to the first service link. This method optimizes the user terminal's perception, that is, after the user terminal initiates a request, it starts transmitting the target content data as soon as possible through the established service link, and then optimizes the transmission quality by switching service links.

[0031] The solution provided by the embodiments of this application can establish a first service link between the MEC and the cloud center based on the content data request of the user terminal, and establish a second service link between the user terminal and the cloud center. Then, the transmission quality of these two service links can be predicted, and the user terminal can be controlled to obtain the target content data from the service link with higher transmission quality through UPF, making full use of link transmission resources and effectively improving the efficiency of content data acquisition for communication users.

[0032] Based on the solution provided in the above embodiments, optionally, as shown in FIG2, before step S12 above, that is, before establishing a first service link between the MEC and the cloud center based on the content data request, and before instructing the user terminal to establish a second service link between the user terminal and the cloud center based on the content data request, the solution further includes: S21: Parse the content data request to obtain the address information in the indication information.

[0033] In the solution provided in this application embodiment, after receiving a content data request, the MEC can obtain the address information (URL or IP) of the content data requested by the terminal according to the DNS (Domain Name System) address resolution module or DPI (Deep Packet Inspect) and other methods.

[0034] See Figure 2bThe process shown begins with the user terminal sending a content data request to the MEC. The MEC can then resolve the request using methods such as DNS or DPI to obtain the indication information carried in the content data request, which may be, for example, the resource address of the target content data.

[0035] S22: Query the cached data of the MEC to see if there is any data that matches the content data request.

[0036] In this step, MEC checks the cached data to see if there is any data matching the requested content data. If so, it can directly perform traffic splitting, execute the local content data service, and provide the cached target content data to the user terminal.

[0037] S23: If the request is not successful, a first service link is established between the MEC and the cloud center based on the content data request, and the user terminal is instructed to establish a second service link between the user terminal and the cloud center based on the content data request.

[0038] If the target content data is not found in the cached data, the request needs to be forwarded to the cloud center to establish two business links to obtain the target content data from the cloud center.

[0039] On one hand, the MEC establishes the first business link with the cloud center. Based on the resolved content data address information, the MEC triggers content data requests for the same address through its own data caching mechanism. Then, the local UPF on the MEC side, through the routing capabilities of the core network UPF, establishes the business link between the MEC data caching module and the cloud center.

[0040] On the other hand, MEC instructs user terminals to establish a second service link with the cloud center. The terminal side establishes a service data link with the cloud center through the routing capability of UPF (local and core network), which enables data flow.

[0041] Subsequently, the link efficiency comparison module can determine which of the two service links has higher transmission quality, thereby maintaining the target service link with higher transmission quality. This ensures that the user terminal uses the better service link to obtain the target content data, effectively improving the transmission efficiency of the target content data.

[0042] Based on the solutions provided in the above embodiments, optionally, such as Figure 3 As shown, in step S12 above, establishing a first business link between the MEC and the cloud center based on the content data request includes: S31: Based on the address information, send the content data request to the cloud center through the local UPF device and the core network UPF device to establish a first service link between the MEC and the cloud center.

[0043] The solution provided in this application embodiment, in the application scenario of local traffic splitting of service data in MEC, when the user's content data request does not hit the cached resources of MEC, is sent from the local UPF device to the core network UPF device via N9, and then routed to the cloud center, thereby realizing the establishment of a service link between the terminal and the cloud center.

[0044] On the other hand, based on the content address information of the terminal obtained by MEC, MEC triggers a business data request to the cloud center, establishing two links for the same content data and achieving dual-path linking for content data acquisition. During the data flow, the link is switched or released based on a comparison of the efficiency of content resources acquired by the user and the MEC caching module. This rational allocation of resource acquisition methods ensures the timeliness of content data acquisition and business quality, fully leveraging the network information development capabilities of edge computing.

[0045] Based on the solutions provided in the above embodiments, optionally, such as Figure 4 As shown, in step S13 above, predicting the transmission quality of the target content data through the first service link and the second service link to determine the target service link with higher transmission quality between the first service link and the second service link includes: S41: Obtain the historical transmission parameters of the service link to be predicted, wherein the service link to be predicted is the first service link or the second service link.

[0046] The solution provided in this application embodiment can be implemented through a link transmission efficiency comparison module, which can determine the transmission efficiency of the first service link and the second service link respectively, so as to quantitatively evaluate the transmission quality of these two service links.

[0047] The historical transmission parameters obtained in this step can be the transmission parameters of the service link to be predicted within a recent period. For example, in one application scenario, after a second service link is established between the user terminal and the cloud center, the transmission of target content data can begin immediately. In this step, historical transmission data can be obtained based on the data interaction actions that are already taking place between the user terminal and the cloud center, combined with the wired portion (cloud center to local UPF) and the wireless portion (gNB to terminal).

[0048] In one application scenario, the first business link established between MEC and cloud center may include a wired transmission portion, for which historical transmission data is obtained.

[0049] The historical transmission data may include information such as maximum bandwidth (kbps) and its utilization, link data transmission loop delay (ms), data transmission traffic (Byte), data download time (ms), number of s data packets (pkt), and data packet transmission quality (pktretrans), which are used to calculate transmission efficiency in subsequent steps.

[0050] S42: Based on the historical transmission parameters and the corresponding feature weights, the transmission efficiency of the service link to be predicted within the target time period is predicted using a two-level exponential smoothing algorithm, wherein the feature weights are determined based on the periodicity and / or volatility of the corresponding historical transmission parameters.

[0051] In this step, the transmission efficiency is calculated based on the historical transmission parameters obtained in the previous steps. Optionally, the link transmission efficiency can be calculated as follows: EFF k =(Byte*8 / (ms*1.024)) / bw-(pktretrans / pkt) Here, the subscript 'k' represents the identifier of a specific service request. Taking video services as an example, a service request consists of an M3U8 index file and the corresponding video TS file.

[0052] Subsequently, by combining historical transmission data (MEC and radio side data caching and distribution efficiency) and the network information opening function of the MEC side, the target switching link transmission efficiency prediction is realized, and the final decision is made on the service link.

[0053] In this step, a two-stage exponential smoothing algorithm is used to predict the transmission efficiency of the service link to be predicted, as detailed below: EFF K+1 (2) =a*(a*EFF k +(1-a)EFF k -1)+(1-a)EFF k-1 (2) Where 'a' is a weighting coefficient, referred to as the feature weight in this example, it is determined based on the periodicity and / or volatility of the time series in the historical transmission data. The actual rule for determining the feature weight can be flexibly set according to requirements. For example, the feature weight can be determined in the following way: When the time series shows a stable horizontal trend, α should be taken as a small value, such as 0.1 to 0.2; When the time series fluctuates significantly and exhibits strong periodicity, α should be taken as a small to medium value, such as 0.3 to 0.5. When the time series fluctuates and exhibits strong periodicity, α should be taken as a medium to large value, such as 0.6 to 0.8; When the time series is stable and without fluctuations, α should be a large value, such as 0.9 to 1.0.

[0054] The following example illustrates this solution. In this example, the transmission path from MEC to the cloud center is relatively stable, meaning the time series shows a stable horizontal trend. However, the user location and service behavior from gNB to the terminal side exhibit some randomness, meaning the amplitude of variation is relatively large.

[0055] Suppose there are three consecutive business requests (K=1,2,3), each involving data transmission from the cloud center to the terminal. Below are some key parameter values ​​for these three requests: For K=1: EFF1=(100000 / 8 / (200 / 1.024)) / 5000-(5 / 500) For K=2: EFF2=(110000 / 8 / (210 / 1.024)) / 4800-(8 / 550) For K=3: EFF3=(120000 / 8 / (220 / 1.024)) / 4700-(12 / 60) By calculating these values, the actual link transmission efficiency for each service request can be obtained. The link transmission efficiency for the fourth service request (K+1) is predicted using a two-stage exponential smoothing algorithm. Assuming the weighting coefficient α is an intermediate value, such as α = 0.4 (this value is chosen because of the significant fluctuations between the gNB and the terminal), then: EFF4(2)=a*(a*EFF3+(1-a)EFF2)+(1-a)EFF3(2) EFF3(2) is the value predicted in the previous prediction. Assuming that the initial prediction value is the same as the actual value, the above formula can be used to predict the next link transmission efficiency. The process can continue, thereby continuously updating the prediction model and adapting to changes in network conditions, thus improving the effectiveness of transmission quality prediction.

[0056] S43: The service link with higher transmission efficiency between the first service link and the second service link is identified as the target service link.

[0057] In this step, based on the transmission efficiency predicted in the previous steps, the service link with higher transmission efficiency is identified as the target service link. In subsequent steps, the terminal is controlled to obtain the target content data through the target service link with significantly higher transmission efficiency, thus ensuring the transmission quality of the target content data.

[0058] Based on the solutions provided in the above embodiments, optionally, such as Figure 5aAs shown, in step S42 above, predicting the transmission efficiency of the service link to be predicted within the target time period using a two-stage exponential smoothing algorithm based on the historical transmission parameters and corresponding feature weights includes: S51: Perform characteristic detection on the historical transmission parameters to obtain multiple characteristic indicators, including fluctuation characteristic indicators and / or periodic characteristic indicators.

[0059] In practical applications, the content resources that MEC needs to cache vary, and the weighting coefficients often need to be set accordingly based on the characteristics of data packets for different services.

[0060] In this embodiment of the application, in terms of data feature recognition, deep intelligent recognition of flow features can be established, features of various business data objects can be extracted, and samples can be mapped into high-dimensional feature vectors.

[0061] Taking streaming media services as an example, the data unit for traffic problem processing based on machine learning is a stream (e.g., a sequence of data packets with the same five-tuple). For a stream sample x, information such as average packet length avg_pack and minimum packet interval min_timegap can be extracted, and this information is mapped to a fixed dimension of a vector, thereby constructing a feature vector x=[avg_pack, min_timegap, ... ...] for the stream, as shown in the table below:

[0062] In the solution provided in this application embodiment, a feature recognition model is constructed based on MEC historical cache data and combined with a business vector matrix to support the output of feature vectors for different business types. This model can effectively distinguish between MEC cached video content, enterprise project software content, data streams from different business lines, and other random business information, and supports effective classification by weighted coefficients.

[0063] The business tag can be defined as a parameter. See [link / reference] for this example. Figure 5b The diagram illustrates matrix operations. In matrix A (t1), c... i The first column of matrix B (t2) represents different network elements and the second column represents various data feature values. Matrix A and matrix B are summed and iteratively calculated according to the time sequence to support the periodic row classification results output of the weighted coefficients.

[0064] In matrix t1, ci1 in the first column represents different types of business identifiers, and kpi01 in the subsequent columns is located in the same row. t1 kpi01 t2 ...kpi01 tnThis refers to the feature vector values ​​of the same feature value across different time slices. For example, the feature vector values ​​of the packet size across slices every 500ms. Similarly, kpi02... t1 kpi02 t2 For example, the feature vector values ​​of the packet interval on different time slices.

[0065] In matrix t2, mme1 represents the identifier of a certain end-to-end network element for the same type of service. The subsequent columns contain kpi01 located in the same row. t1 kpi01 t2 This refers to the feature vector values ​​of the same feature value (such as packet size) at different time slices (each slice is 500ms). Similarly, kpi02 t1 kpi02 t2. These are the feature vector values ​​(e.g., packet intervals) on different time slices.

[0066] In this embodiment, the service identifier can represent the overall feature quantization value of the service on the end-to-end link, and the network element identifier is the feature quantization value of the service on different network element nodes. In this scheme, the offset calculation of the same service on different network elements is realized through matrix addition operation, supporting feature identification on the link from edge MEC to core network to base station to terminal, and realizing the differentiation of service and packet feature status based on the feature calculation results.

[0067] S52: Predict multiple characteristic indicators by using a preset algorithm model for matching characteristic indicators, and obtain multiple predicted characteristic indicators.

[0068] In this step, multiple dimensions such as time, space, network elements, and terminals can be considered, along with three types of data fluctuation characteristics: strong periodicity, weak periodicity, and interval-based fluctuations. DNN, LightGBM, and HistoryMean algorithms can be integrated, and the Transformer algorithm can be used for weight optimization to achieve link quality prediction across different dimensions. The parameter feature detection process is as follows: Figure 5c As shown.

[0069] This solution detects the volatility and stationarity of the collected time-series data. Based on the detection results, the indicators are categorized into non-volatile and volatile indicators. Volatile indicators are further classified into aperiodic, strongly periodic, and weakly periodic indicators based on characteristics such as autocorrelation and periodicity. Building upon the automatic identification and classification of indicators, different machine learning and deep learning algorithms are employed for indicators with different characteristics to predict link quality indicators.

[0070] S53: Determine the feature weights corresponding to the multiple predictive characteristic indicators respectively through a multi-head self-attention transformation model.

[0071] See Figure 5dIn step S52, the historical time series data is input into the corresponding model algorithm according to the detection results to perform prediction, resulting in multiple multinomial prediction feature indicators. For example, prediction result 1 is obtained through the DNN algorithm, prediction result 2 is obtained through the LightGBM algorithm, and prediction result 3 is obtained through the History Mean algorithm. In this step, the multiple prediction feature indicators are input into the multi-head self-attention transform model (Transformer model), which includes an encoder and a decoder, with the structure as follows: Figure 5d As shown, the prediction results of the above algorithm are corrected by a multi-investment attention transformation model, and the prediction results are fused and optimized to obtain the optimized prediction results. Specifically, for different prediction algorithms with different periodic characteristics, in order to improve the dynamic prediction of link quality, a multi-prediction algorithm ensemble model is used, combined with the Transformer algorithm, to optimize the weights and achieve the optimal prediction result output.

[0072] S54: Based on the multiple predictive characteristic indicators and their corresponding feature weights, the transmission efficiency of the service link to be predicted within the target time period is predicted using a two-level exponential smoothing algorithm.

[0073] In this step, the optimized feature weights obtained from the aforementioned algorithm model are used to predict the transmission efficiency of the service link to be predicted within the target time period using a two-stage exponential smoothing algorithm. This scheme, considering the volatility and stationarity of time-series data, employs an appropriate algorithm model to perform predictions, fully utilizing the model's advantages to effectively predict multi-dimensional features. Furthermore, a multi-head self-attention transformation model is used to achieve optimization and integration, improving the effectiveness of feature weights and optimizing the quality of the predicted transmission efficiency.

[0074] Based on the solutions provided in the above embodiments, optionally, such as Figure 6a As shown, in step S12 above, instructing the user terminal to establish a second service link between the user terminal and the cloud center based on the content data request includes: S61: Instruct the user terminal to establish a second business link between the user terminal and the cloud center based on the content data request, and obtain the target content data through the second business link.

[0075] In this step, MEC instructs the user terminal to establish a second service link between the user terminal and the cloud center. For example, this could include instructing the user terminal to route content data requests to the cloud center via both the local UPF and the central UPF, thereby establishing a second service link between the cloud center and the user terminal.

[0076] After establishing the second service link, there's no need to wait for the first service link to be established or for transmission quality comparison results to be obtained. The user terminal can directly initiate the acquisition of target content data through the second service link. In this way, the user terminal can quickly perceive the content data transmission after initiating a content data request, optimizing the user terminal experience.

[0077] Where the target service link is the first service link, in step S14 above, controlling the user terminal to obtain at least a portion of the target content data through the target service link via the local user plane function (UPF) device includes: S62: Monitor the progress of the user terminal in acquiring the target content data through the second service link via the local UPF device, and when the data acquired by the user terminal and the data cached by the MEC meet the preset synchronization rules, control the user terminal to switch from the second service link to the first service link to continue acquiring the target content data.

[0078] In the solution provided in this application embodiment, if the target service link is the first service link, it is necessary to control the user terminal to switch links and use the first service link with better transmission quality to continue the subsequent transmission of the target service link.

[0079] In this step, the progress of the target content data already acquired by the user terminal is monitored by the local UPF device, as well as the progress of the cached data of MEC, so as to select an appropriate time to switch links.

[0080] Among them, see Figure 6b As shown in the process, when a user terminal obtains content data through a service link, the UPF can use the application layer interaction information recognition capability of DPI to track the message information interaction of the terminal link and obtain the terminal's content data request message information. Based on the routing function of the local UPF, the busy / idle state monitoring of the transport layer link is realized, and the UPF pushes the content data index to the MEC side. The MEC can obtain the user terminal's subsequent content data in advance based on the content data message interaction. Specifically, the MEC requests content data after increasing the message step size to obtain the content data that the user terminal will obtain in the future. The UPF monitors the content data process on the terminal side and performs content data synchronization detection on the MEC side. When the data is consistent with the data obtained in advance by the MEC side, the local UPF switches the link from the second service link where the terminal is located to the first service link where the MEC is located, so that the content data continues to flow, and the MEC distributes the subsequent content data transmitted through the first service link to the user terminal to enable the user terminal to obtain the complete target content data. In addition, the UPF can also initiate a terminal-side link release request to the cloud center to avoid the user terminal occupying additional link transmission resources.

[0081] The solution provided in this application embodiment can acquire application layer information in the terminal-side link through the DPI function, such as content data address information and media format, effectively enabling the terminal to monitor the progress of acquiring target content data. Furthermore, the UPF can push message information such as content data address to the MEC's ​​cache module, allowing the MEC to request content data in three pre-defined steps (the number can be flexibly set) based on the message information. For example, adding an ID to the existing ID of the M3U8 index file enables video indexing and data acquisition. Then, when the content data acquisition process of the user terminal is consistent with the data carried in the message initially acquired by the MEC, it is determined that a preset synchronization rule is met, and the UPF switches the link, enabling the MEC to distribute subsequent data to the user terminal. Additionally, the UPF proactively initiates a link release request before the terminal switches.

[0082] In the solution provided in this application, regarding link construction, when cached data is not found, the link between the MEC side and the cloud center is established and kept active, effectively enabling flexible link switching and efficient data distribution. Regarding link optimization, this solution, based on the historical data distribution efficiency of the MEC and wireless side links and combined with the MEC's ​​wireless network development capabilities, effectively predicts link transmission efficiency using a quadratic exponential smoothing algorithm, thus predicting the continuous service distribution capability of the target switching link. Regarding user terminal perception, this solution, based on the routing function of the local UPF and the acquisition of content data in different processes across dual links, effectively enhances the MEC's ​​traffic distribution function. Simultaneously, the priority establishment of the terminal-side link ensures the real-time nature of user service perception. This solution fully leverages the superior speed of MEC-side data caching, enabling advance acquisition of content data. When synchronized with terminal data, data is acquired by the MEC side and distributed to the user. It effectively combines the MEC's ​​network information development capabilities to achieve dynamic matching of content data and network status, ensuring user perception.

[0083] To address the problems existing in related technologies, this application also provides a content data acquisition device 70 based on dual links, such as... Figure 7 As shown, the device applied to the edge computing (MEC) includes: The receiving module 71 receives a content data request from a user terminal, wherein the content data request carries indication information of the target content data requested by the user terminal. Establishment module 72, based on the content data request, establishes a first service link between the MEC and the cloud center, and instructs the user terminal to establish a second service link between the user terminal and the cloud center based on the content data request; Prediction module 73 predicts the transmission quality of the first service link and the second service link to the target content data, so as to determine the target service link with high transmission quality among the first service link and the second service link; The control module 74 controls the user terminal to obtain at least a portion of the target content data through the target service link via the local user plane function (UPF) device.

[0084] The apparatus provided in this application first receives a content data request from a user terminal, the content data request carrying indication information of the target content data requested by the user terminal; then, a first service link is established between the MEC and the cloud center based on the content data request, and the user terminal is instructed to establish a second service link between the user terminal and the cloud center based on the content data request; next, the transmission quality of the target content data through the first and second service links is predicted to determine the target service link with higher transmission quality; subsequently, the user terminal is controlled to obtain at least a portion of the target content data through the target service link via the local user plane function (UPF) device. The solution provided in this application can establish a first service link between the MEC and the cloud center based on the content data request of the user terminal, and establish a second service link between the user terminal and the cloud center, thereby predicting the transmission quality of these two service links, and controlling the user terminal to obtain the target content data from the service link with higher transmission quality through the UPF, fully utilizing link transmission resources and effectively improving the efficiency of content data acquisition for communication users.

[0085] In this application, the modules in the apparatus provided can also implement the method steps provided in the method embodiments. Alternatively, the apparatus provided in this application may include other modules besides those described above to implement the method steps provided in the method embodiments. Furthermore, the apparatus provided in this application can achieve the technical effects achievable by the method embodiments.

[0086] Preferably, this application embodiment also provides an electronic device, including a processor, a memory, and a computer program stored in the memory and executable on the processor. When the computer program is executed by the processor, it implements the various processes of the above-described embodiment of a dual-link content data acquisition method and achieves the same technical effect. To avoid repetition, it will not be described again here.

[0087] This application also provides a computer-readable storage medium storing a computer program. When executed by a processor, this computer program implements the various processes of the above-described dual-link content data acquisition method embodiment and achieves the same technical effect. To avoid repetition, it will not be described again here. The computer-readable storage medium may be a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.

[0088] This application also provides a computer program product, which includes a non-transitory computer-readable storage medium storing a computer program. The computer program is operable to cause a computer to perform some or all of the steps of the above-described embodiment of a dual-link content data acquisition method, and can achieve the same technical effect. To avoid repetition, it will not be described again here.

[0089] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0090] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in one or more blocks of the flowchart illustrations and / or one or more blocks of the block diagrams.

[0091] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means that implement the functions specified in one or more flowcharts and / or one or more block diagrams.

[0092] These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer-implemented process, such that the instructions, which execute on the computer or other programmable apparatus, provide steps for implementing the functions specified in one or more flowcharts and / or one or more block diagrams.

[0093] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0094] Memory may include non-persistent storage in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM. Memory is an example of computer-readable media.

[0095] Computer-readable media includes both permanent and non-permanent, removable and non-removable media that can store information using any method or technology. Information can be computer-readable instructions, data structures, modules of programs, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic magnetic disk storage or other magnetic storage devices, or any other non-transferable medium that can be used to store information accessible by a computing device. As defined herein, computer-readable media does not include transient computer-readable media, such as modulated data signals and carrier waves.

[0096] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0097] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0098] The above description is merely an embodiment of this application and is not intended to limit this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principle of this application should be included within the scope of the claims of this application.

Claims

1. A method for acquiring content data based on dual links, characterized by, Applied to an edge computing device MEC, comprising: receiving a content data request of a user terminal, the content data request carrying indication information of target content data requested by the user terminal to obtain; establishing a first service link between the MEC and a cloud center based on the content data request, and instructing the user terminal to establish a second service link between the user terminal and the cloud center based on the content data request; predicting the transmission quality of the target content data of the first service link and the second service link to determine the target service link with high transmission quality in the first service link and the second service link; controlling the user terminal to obtain at least part of the target content data through the target service link through a local user plane function UPF device.

2. The method of claim 1, wherein, Before establishing a first service link between the MEC and a cloud center based on the content data request, and instructing the user terminal to establish a second service link between the user terminal and the cloud center based on the content data request, it also includes: parsing the content data request to obtain address information in the indication information; querying whether there is data that hits the content data request in the cache data of the MEC; if not, establish a first service link between the MEC and a cloud center based on the content data request, and instruct the user terminal to establish a second service link between the user terminal and the cloud center based on the content data request.

3. The method of claim 2, wherein, Establishing a first service link between the MEC and a cloud center based on the content data request includes: based on the address information, sending the content data request to the cloud center through a local UPF device and a core network UPF device to establish a first service link between the MEC and the cloud center.

4. The method of claim 1, wherein, Predicting the transmission quality of the target content data of the first service link and the second service link to determine the target service link with high transmission quality in the first service link and the second service link includes: obtain historical transmission parameters of the to-be-predicted service link, the to-be-predicted service link being the first service link or the second service link; predicting the transmission efficiency of the to-be-predicted service link in the target period based on the historical transmission parameters and the corresponding feature weight through a two-level exponential smoothing algorithm, wherein the feature weight is determined based on the periodicity and / or volatility of the corresponding historical transmission parameters; determining the service link with high transmission efficiency in the first service link and the second service link as the target service link.

5. The method of claim 4, wherein, Predicting the transmission efficiency of the to-be-predicted service link in the target period based on the historical transmission parameters and the corresponding feature weight through a two-level exponential smoothing algorithm includes: performing characteristic detection on the historical transmission parameters to obtain a plurality of characteristic indexes, the characteristic indexes including fluctuation characteristic indexes and / or periodic characteristic indexes; performing prediction on a plurality of characteristic indexes through a preset algorithm model matched with the characteristic indexes to obtain a plurality of predicted characteristic indexes; determining the feature weight corresponding to the plurality of predicted characteristic indexes through a multi-head self-attention transformation model; predicting, based on the multiple predicted characteristic indicators and respectively corresponding characteristic weights, a transmission efficiency of the to-be-predicted service link in a target period by a second exponential smoothing algorithm.

6. The method according to any one of claims 1 to 5, wherein, indicating the user terminal to establish a second service link between the user terminal and a cloud center based on the content data request, comprising: indicating the user terminal to establish a second service link between the user terminal and a cloud center based on the content data request and acquiring the target content data through the second service link; wherein, in a case that the target service link is the first service link, the controlling the user terminal to acquire at least part of the target content data through the target service link by a local user plane function (UPF) device, comprising: monitoring, by the local UPF device, a progress of the user terminal in acquiring the target content data through the second service link, and when the acquired data of the user terminal and the cached data of the MEC satisfy a preset synchronization rule, controlling the user terminal to switch from the second service link to the first service link to continue acquiring the target content data.

7. A dual-link based content data acquisition apparatus, characterized by comprising: applied to an edge computing device (MEC), comprising: a receiving module configured to receive a content data request of a user terminal, the content data request carrying indication information of target content data requested to be acquired by the user terminal; a establishing module configured to establish a first service link between the MEC and a cloud center based on the content data request, and instructing the user terminal to establish a second service link between the user terminal and the cloud center based on the content data request; a predicting module configured to predict transmission qualities of the first service link and the second service link on the target content data to determine a target service link with a high transmission quality from the first service link and the second service link; a controlling module configured to control the user terminal to acquire at least part of the target content data through the target service link by a local user plane function (UPF) device.

8. An electronic device, comprising: comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, the computer program being executed by the processor to implement steps of the method according to any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that, The computer program is stored on the computer readable storage medium and is executable by the processor to implement steps of the method according to any one of claims 1 to 6.

10. A computer program product, characterised in that, The computer program product comprises a non-transitory computer readable storage medium storing a computer program operable to cause a computer to implement steps of the method according to any one of claims 1 to 6.

Images