Data processing method and device, equipment and storage medium

Abstract

The application discloses a data processing method and device, equipment and storage medium, and relates to the technical field of data processing. The method is based on the structural parameters of a network topology structure, node state information, service traffic data entering a communication network, network state parameters and other factors, and predicts network capacity information of the communication network through a network capacity prediction model, which is beneficial to accurately generating a dynamic mapping relationship between service traffic data processed by the communication network in a first time period and load states of each node in the communication network in the first time period, improving the adaptability of dynamic complex network traffic changes and the accuracy of calculating long-term communication network capacity in a network dynamic complexity, and mapping the first network capacity information to the network topology structure, so that the load states of each node and a link connecting at least two nodes in the first time period can be intuitively and clearly viewed through a network capacity distribution visual map.

Description

Technical Field

[0001] This application belongs to the field of data processing technology, and in particular relates to a data processing method, apparatus, device and storage medium. Background Technology

[0002] Communication network capacity refers to the ability of communication networks, such as transmission networks and core networks, to handle network traffic. In related technologies, communication network capacity can be obtained through static analysis based on network flow models, short-term assessments based on time series analysis, load balancing based on redundancy allocation, or load control based on service priorities. However, the aforementioned methods are poorly adaptable to dynamic and complex network traffic changes, and therefore cannot accurately calculate long-term communication network capacity under dynamic and complex network conditions. Summary of the Invention

[0003] This application provides a data processing method, apparatus, device, and storage medium that can solve the problem in related technologies of the inability to accurately calculate the long-term communication network capacity under dynamic and complex network conditions.

[0004] In a first aspect, embodiments of this application provide a data processing method, which may include:

[0005] Obtain network information of the communication network. The communication network has a network topology structure, which includes N nodes and a first link connecting every two nodes. The network information includes the structural parameters of the network topology structure, the node status information of the N nodes, the service traffic data entering the communication network, and the network status parameters. The network status parameters include the parameters generated during the process of the communication network processing the service traffic data.

[0006] Based on network information, the first network capacity information of the communication network in the first time period is calculated using a network capacity prediction model.

[0007] The first network capacity information is mapped onto the network topology to obtain a network capacity distribution visualization map. The network capacity distribution visualization map includes P nodes out of N nodes and a second link for connecting every two nodes out of P nodes. The network capacity distribution visualization map is used to characterize the first load state of each of the P nodes in the first time period and the second load state of the second link in the first time period, where P∈[1,N].

[0008] Secondly, embodiments of this application provide a data processing apparatus, which may include:

[0009] The acquisition module is used to acquire network information of the communication network. The communication network has a network topology structure, which includes N nodes and a first link connecting every two nodes. The network information includes the structural parameters of the network topology structure, the node status information of the N nodes, the service traffic data entering the communication network, and the network status parameters. The network status parameters include the parameters generated during the process of the communication network processing the service traffic data.

[0010] The calculation module is used to calculate the first network capacity information of the communication network in the first time period based on network information and through the network capacity prediction model;

[0011] The mapping module is used to map the first network capacity information to the network topology to obtain a network capacity distribution visualization map. The network capacity distribution visualization map includes P nodes out of N nodes and a second link for connecting every two nodes out of P nodes. The network capacity distribution visualization map is used to characterize the first load state of each node out of P nodes in the first time period and the second load state of the second link in the first time period, where P∈[1,N].

[0012] Thirdly, embodiments of this application provide a computer device, which includes: a processor and a memory storing computer program instructions;

[0013] When the processor executes computer program instructions, it implements the data processing method as described in the first aspect.

[0014] Fourthly, embodiments of this application provide a computer storage medium storing computer program instructions, which, when executed by a processor, implement the data processing method as described in the first aspect.

[0015] Fifthly, embodiments of this application provide a chip, which includes a processor and a communication interface. The communication interface and the processor are coupled, and the processor is used to run programs or instructions to implement the data processing method as shown in the first aspect.

[0016] In a sixth aspect, embodiments of this application provide a computer program product stored in a storage medium, which is executed by at least one processor to implement the data processing method as described in the first aspect.

[0017] The data processing method, apparatus, device, and storage medium of this application embodiment acquire network information of a communication network. The communication network has a network topology structure, which includes N nodes and a first link connecting every two nodes among the N nodes. The network information includes structural parameters of the network topology structure, node status information of the N nodes, service traffic data entering the communication network, and network status parameters. The network status parameters include parameters generated during the processing of service traffic data by the communication network. Based on the network information, a first network capacity information of the communication network in a first time period is calculated using a network capacity prediction model. The first network capacity information is mapped onto the network topology structure to obtain a network capacity distribution visualization map. The network capacity distribution visualization map includes P nodes among the N nodes and a second link connecting every two nodes among the P nodes. The network capacity distribution visualization map is used to characterize the first load state of each of the P nodes in the first time period and the second load state of the second link in the first time period. In this way, based on factors such as network topology structural parameters, node status information, incoming traffic data, and network status parameters, a network capacity prediction model can be used to predict the network capacity information of the communication network. This facilitates the accurate generation of a dynamic mapping relationship between the traffic data processed by the communication network in the first time period and the load status of each node in the communication network in the first time period. This improves the adaptability to dynamic and complex network traffic changes, thereby enhancing the accuracy of calculating long-term communication network capacity under dynamic and complex network conditions. This enables proactive and precise monitoring of the communication network capacity information, detection of potential faults, and, by mapping the first network capacity information onto the network topology, a visualized network capacity distribution map can reflect the load status of each node and the link connecting at least two nodes in the first time period. This allows users to intuitively and clearly view the traffic data processed by each node and link in the first time period. Attached Figure Description

[0018] To more clearly illustrate the technical solutions of the embodiments of this application, the accompanying drawings used in the embodiments of this application will be briefly introduced below. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0019] Figure 1 This is a schematic diagram of the structure of a data processing system provided in an embodiment of this application;

[0020] Figure 2 A flowchart illustrating a data processing method provided in an embodiment of this application;

[0021] Figure 3A flowchart illustrating a data processing method provided in an embodiment of this application;

[0022] Figure 4 This is a schematic diagram of the structure of a data processing apparatus provided in one embodiment of this application;

[0023] Figure 5 This is a schematic diagram of the structure of a computer device provided in one embodiment of this application. Detailed Implementation

[0024] The features and exemplary embodiments of various aspects of this application will be described in detail below. To make the objectives, technical solutions, and advantages of this application clearer, the application will be further described in detail below with reference to the accompanying drawings and specific embodiments. It should be understood that the specific embodiments described herein are only intended to explain this application and not to limit it. For those skilled in the art, this application can be implemented without some of these specific details. The following description of the embodiments is merely to provide a better understanding of this application by illustrating examples.

[0025] It should be noted that, in this document, relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, 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 limitations, an element defined by the phrase "comprising..." does not exclude the presence of additional identical elements in the process, method, article, or apparatus that includes the element.

[0026] In related technologies, communication network capacity can be obtained through static analysis based on network flow models, short-term assessment based on time series analysis, load balancing based on redundancy allocation, or load control based on service priorities. Specifically, static analysis based on network flow models establishes a deterministic network flow model based on the network topology and performs static analysis based on the current network load to predict network congestion nodes. However, this method is poorly adaptable to dynamic and complex service traffic and cannot achieve accurate prediction of network capacity information. Short-term prediction based on time series analysis collects historical network traffic data and uses time series analysis to predict short-term network traffic, achieving a short-term assessment of network capacity. However, this method struggles to assess long-term network stability and cannot guide network planning. Load balancing based on redundancy allocation sets backup resources at key network nodes, enabling backup resources to distribute load when the primary resource is overloaded. While this can improve network carrying capacity, resource utilization is low. Load control based on service priorities implements priority control for different service traffic, restricting low-priority services to ensure high-priority services during network congestion, but it cannot achieve global network resource optimization.

[0027] It is evident that the aforementioned methods are poorly adaptable to dynamic and complex network traffic changes. Therefore, they cannot accurately calculate long-term communication network capacity under dynamic and complex network conditions, and cannot achieve proactive and precise network capacity monitoring.

[0028] To address the aforementioned technical problems, embodiments of this application provide a data processing method, apparatus, computer equipment, and storage medium.

[0029] Therefore, in order to better illustrate the content of the embodiments of this application, the following will be combined with... Figures 1 to 5 The following describes a data processing system, method, apparatus, and computer device provided in the embodiments of this application.

[0030] Figure 1 This is a schematic diagram of the structure of a data processing system provided in an embodiment of this application.

[0031] like Figure 1 As shown, the data processing system 10 may include a communication network topology module 101, a node feature module 102, a service traffic module 103, a machine learning module 104, a prediction and early warning module 105, and a network management interface module 106.

[0032] Specifically, the communication network topology module 101 is used to construct the network topology of the communication network based on information involved in the communication network, such as device information of N physical devices, data transmission relationship information between every two physical devices, connection relationship information between every two physical devices, and service traffic entering the communication network. The network topology includes nodes corresponding to each of the N physical devices in the communication network and a first link connecting every two nodes, where N is an integer greater than 1.

[0033] The node feature module 102 is used to monitor the static attributes and dynamic status parameters of each node in the communication network in real time and construct node status information. The static attributes include at least one of the following: the physical hardware specifications of the node, the processing power of the processor in the node, and the number of fiber optic ports; the dynamic status parameters may include at least one of the following: port traffic, processor utilization in the node, actual memory usage, routing information, cache configuration, packet loss rate, latency, and jitter. In this embodiment, the node is a physical device, such as a mobile phone, computer, or server.

[0034] The service traffic module 103 is used to collect network information and current network information within the communication network during the second time period. It also establishes sample information based on the network information within the second time period. The network information includes structural parameters of the network topology, node status information of N nodes, service traffic data entering the communication network, and network status parameters. The network status parameters include parameters generated during the processing of service traffic data by the communication network.

[0035] Machine learning module 104 is used to train a network capacity prediction model based on a recurrent neural network to model the mapping relationship between traffic and capacity.

[0036] The prediction and early warning module 105, based on the network capacity prediction model output by the machine learning module 104, calculates the first network capacity information of the communication network in the first time period based on the current network information output by the service traffic module 103, and then issues early warning information based on the first network capacity information.

[0037] The network management interface module 106 is used to connect and interact with the network management system to collaboratively complete network status monitoring and scheduling management.

[0038] By combining the aforementioned modules, a data processing system for intelligent communication network capacity can be constructed. This system can predict network capacity information based on various factors, including network topology parameters, node status information, incoming traffic data, and network status parameters. This prediction, achieved through a network capacity forecasting model, facilitates accurate generation of a dynamic mapping between the traffic data processed by the network in the first time period and the load status of each node within the network during that time period. This improves adaptability to dynamic and complex network traffic changes, thereby enhancing the accuracy of long-term communication network capacity calculations under dynamic and complex network conditions. Ultimately, this enables proactive and precise monitoring of network capacity information and the detection of potential faults. Furthermore, by mapping the initial network capacity information onto the network topology, a visualized network capacity distribution map can reflect the load status of each node and the load status of links connecting at least two nodes in the first time period. This allows users to intuitively and clearly view the traffic data processed by each node and link within the first time period.

[0039] Based on this, such as Figure 2 As shown in the figure, this application provides a data processing method.

[0040] Figure 2 This is a flowchart of a data processing method provided in an embodiment of this application.

[0041] The data processing method provided in this application can be applied to computer devices, and the data processing method may specifically include the following steps:

[0042] Step 210: Obtain network information of the communication network. The communication network has a network topology structure, which includes N nodes and a first link connecting every two nodes among the N nodes. The network information includes structural parameters of the network topology structure, node status information of the N nodes, service traffic data entering the communication network, and network status parameters. The network status parameters include parameters generated during the processing of service traffic data by the communication network. Step 220: Calculate the first network capacity information of the communication network in the first time period using a network capacity prediction model based on the network information. Step 230: Map the first network capacity information onto the network topology structure to obtain a network capacity distribution visualization map. The network capacity distribution visualization map includes P nodes among the N nodes and a second link connecting every two nodes among the P nodes. The network capacity distribution visualization map is used to characterize the first load state of each node among the P nodes in the first time period and the second load state of the second link in the first time period, where P∈[1,N].

[0043] In this way, based on factors such as network topology structural parameters, node status information, incoming traffic data, and network status parameters, a network capacity prediction model can be used to predict the network capacity information of the communication network. This facilitates the accurate generation of a dynamic mapping relationship between the traffic data processed by the communication network in the first time period and the load status of each node in the communication network in the first time period. This improves the adaptability to dynamic and complex network traffic changes, thereby enhancing the accuracy of calculating long-term communication network capacity under dynamic and complex network conditions. This enables proactive and precise monitoring of the communication network capacity information, detection of potential faults, and, by mapping the first network capacity information onto the network topology, a visualized network capacity distribution map can reflect the load status of each node and the link connecting at least two nodes in the first time period. This allows users to intuitively and clearly view the traffic data processed by each node and link in the first time period.

[0044] The steps described above are explained in detail below.

[0045] First, regarding step 210, in some embodiments of this application, the structural parameters include at least one of the following: the node type of each of the N nodes, and the node identifier of the first link-connected node; the node status information includes at least one of the following: the port traffic of each of the N nodes, and the service routing information of each of the N nodes; the network status parameters include at least one of the following: the service latency for processing service traffic data, and the service packet loss rate for processing service traffic data.

[0046] For example, the process involves acquiring node status information of each node in the network, connection relationship data between every two nodes (such as inter-node link bandwidth, connection matrix, etc.), service traffic data entering the communication network, and network status parameters); based on the node status information, connection relationship data between every two nodes (such as inter-node link bandwidth, connection matrix, service traffic data entering the communication network, and network status parameters), a digital model of the network is constructed using graph theory representation techniques, resulting in a communication network with a network topology; identifying nodes: key nodes, such as regional nodes and core landing points, are identified in the network graph, and their type and location information is obtained; identifying links: the connection links between nodes are identified, and parameters such as link length, bandwidth, and latency are obtained; storing the network model: the constructed network topology graph is stored as data for easy subsequent retrieval.

[0047] Next, regarding step 220, in some embodiments of this application, before step 220, it is necessary to obtain the trained network capacity prediction model, which can be obtained through the following steps 2401 and 2402.

[0048] Step 2401: Obtain sample information. The sample information includes sample network information for each of the M sample communication networks. Each sample communication network has a sample network topology. The sample network topology includes K sample nodes and a first sample link connecting every two sample nodes among the K sample nodes. The sample network information includes sample structure parameters of the sample network topology, sample node information of the K sample nodes, sample service traffic data entering the sample communication network, and sample network status parameters. The sample network status parameters include parameters generated during the processing of sample service traffic data by the sample communication network. M is an integer greater than 1, and K∈[1,M].

[0049] Specifically, step 2401 may include steps 24011 to 24014, as shown below.

[0050] Step 24011: Based on the connection information of each sample node in each sample communication network, mark the sample nodes and the sample links used to connect every two sample nodes in the sample network topology diagram to obtain the sample structure parameters of the sample network topology.

[0051] Step 24012: Select K sample nodes from the sample network topology and obtain the node information of each sample node in the K sample nodes; determine the node information of each sample node in the K sample nodes as the sample node information of the K sample nodes.

[0052] Step 24013: Obtain traffic parameters through the traffic detector and determine the sample service type based on the traffic parameters; generate sample service traffic data based on the sample service information and traffic parameters related to the sample service type.

[0053] Step 24014: Monitor the parameters during the process of the sample communication network processing the sample service traffic data, and determine the parameters generated during the process of the sample communication network processing the sample service traffic data as the sample network status parameters.

[0054] Step 2402: Based on the sample information, train the sample network capacity prediction model until the preset training conditions are met to obtain the network capacity prediction model.

[0055] For example, the process involves: determining the model algorithm: for network traffic time-series data, selecting a recurrent neural network algorithm capable of handling time-dependent data, such as LSTM or GRU; constructing the network structure: designing a neural network structure with input layers, multiple hidden layers, and output layers, and determining the number of nodes in each layer; determining the training data: integrating the network topology model, node state data, and traffic data into samples as model training data; setting model parameters: determining the parameters of the network layers, such as weight matrices and threshold function parameters, and setting training hyperparameters such as the number of iterations and their size; and training the model: using the training data, training the network model through the backpropagation algorithm, minimizing the loss function, and obtaining the model parameters.

[0056] Regarding step 220, in some embodiments of this application, step 220 may specifically include the following processes: data preprocessing: cleaning, anomaly removal, normalization, and other preprocessing of the received data; constructing input vectors: organizing the processed data into the input vector format defined by the machine learning model; model initialization: loading the pre-trained network capacity prediction model and initializing the model's weights and parameters; performing model prediction: inputting the constructed input vector into the initialized model, running the model's forward computation process, and obtaining the output vector; parsing capacity output: parsing the output vector and extracting the predicted network capacity information, including the expected load status of each node and link.

[0057] It should be noted that the sample information may be updated before or after step 2402. Therefore, the data processing method may also include:

[0058] With updated sample information, the network capacity prediction model is trained based on the updated sample information until preset training conditions are met, resulting in an optimized network capacity prediction model. This optimized model is used to calculate the network capacity information of the communication network in the fourth time period. The end time of the first time period is earlier than the start time of the fourth time period.

[0059] In some embodiments of this application, if the network capacity prediction model includes at least two candidate network capacity prediction models, a network capacity prediction model that is more compatible with the current communication network can be selected from the at least two candidate network capacity prediction models for calculation. Based on this, before step 220, the data processing method may further include:

[0060] Based on the communication network, a network capacity prediction model is selected from at least two candidate network capacity prediction models, and the number of network capacity prediction models is at least one.

[0061] Specifically, network capacity prediction models can be selected from at least two candidate models using the following two methods: First, if the network capacity prediction model corresponds to a specific network topology (e.g., the first model corresponds to the first network topology, and the second model corresponds to the second network topology), then the network capacity prediction model corresponding to the communication network topology can be obtained. Second, based on the version of the communication network, specifically whether it is an updated model. For example, if at least two network capacity prediction models include a first model before the update and a second model after the update for a particular communication network, then the selection can be based on whether the network topology has been updated.

[0062] Then, regarding step 230, in some embodiments of this application, before or after step 230, it can be determined whether the communication network is overloaded within the first time window based on the first network capacity information. If so, early warning information characterizing the overload of the communication network within the first time period can be generated, so as to generate a network early warning distribution visualization map based on the early warning information and the network capacity distribution visualization map. Based on this, the data processing method may also include steps 2501 and 2502, as detailed below.

[0063] Step 2501: Calculate the network capacity threshold of the communication network based on the network capacity of the communication network and the second network capacity information of the communication network in the second time period, wherein the end time of the second time period is earlier than the start time of the first time period.

[0064] Step 2502: If the first network capacity value corresponding to the first network capacity information is greater than or equal to the network capacity threshold, generate early warning information. The early warning information is used to characterize the overload of the communication network in the first time period.

[0065] Optionally, in this embodiment of the application, the network capacity threshold can be used to further locate the network area in the communication network that is overloaded. Based on this, step 2502 may specifically include steps 25021 and 25022.

[0066] Step 25021: If the first network capacity value corresponding to the first network capacity information is greater than or equal to the network capacity threshold, calculate the degree of overload of the communication network in the first time period based on the difference between the first network capacity value and the network capacity threshold.

[0067] Step 25022: Based on the degree information and the network area of ​​the communication network that is overloaded in the first time period, generate early warning information. The network area includes at least one of the following: a first area where the nodes in the communication network are located, and a second area where the link connecting every two nodes in the communication network is located.

[0068] In some embodiments of this application, since network areas can include multiple types, a process for determining network areas may also be included. Based on this, the first network capacity value includes a first sub-network capacity value and a second sub-network capacity value, and the network capacity threshold includes a first network capacity threshold and a second network capacity threshold. Thus, before step 25022, the data processing method may further include:

[0069] If the capacity value of the first sub-network is greater than or equal to the capacity threshold of the first network, the first region where the node in the communication network is located is determined as the network region.

[0070] If the capacity value of the second sub-network is greater than or equal to the capacity threshold of the second network, the second region where the link connecting every two nodes in the communication network is located is determined as the network region.

[0071] If the capacity value of the first sub-network is greater than or equal to the first network capacity threshold, and the capacity value of the second sub-network is greater than or equal to the second network capacity threshold, the first region and the second region are determined as network regions.

[0072] Furthermore, in some embodiments of this application, since network regions may exist of different types in the embodiments of this application, corresponding response strategies can be generated based on different types of regions in the network. Based on this, after step 25022, the data processing method may also include steps 2601 to 2602.

[0073] Step 2601: Through the network management system, generate information on the factors affecting the network area's services based on the service types carried by the network area.

[0074] Step 2602: Generate a response strategy based on the information on influencing factors.

[0075] Specifically, the influencing factor information in this application embodiment may include at least one of the following: the number of nodes in the communication network is lower than a first preset threshold, the link capacity in the communication network is lower than a second preset threshold, the service routing failure in the network area, the bandwidth resources of the communication network are lower than a third preset threshold, and the service configuration strategy of the communication network is abnormal. Therefore, it can be seen that the influencing factor information in this application embodiment may have different types. Thus, different response strategies can be generated based on different types of influencing factor information. Based on this, step 2602 may specifically include:

[0076] If the influencing factor is that the number of nodes in the communication network is lower than a first preset threshold, a response strategy for increasing the number of nodes in the communication network is generated.

[0077] When the influencing factor information is that the link capacity in the communication network is lower than the second preset threshold, a response strategy for increasing the link bandwidth in the communication network is generated.

[0078] In the event of a service routing failure in a network area, a response strategy is generated to redirect service traffic data processed in the network area to a target network area in the communication network. The target network area does not overlap with the network area.

[0079] When the influencing factor information is that the bandwidth resources of the communication network are lower than the third preset threshold, a response strategy is generated to adjust the network priority for the communication network to process service traffic data.

[0080] When the influencing factor information is that the service configuration policy of the communication network is abnormal, generate response strategies for restricting or isolating the service traffic data processed by nodes in the network area, generate response strategies for updating the load balance of the communication network, or generate response strategies for issuing instructions to manage the processing of service traffic data of the communication network.

[0081] Based on this, in some embodiments of this application, the communication network can be adjusted based on the response strategy, and the network capacity information of the adjusted communication network can be calculated based on the adjusted communication network (here, it can refer to step 230 to recalculate whether the adjusted communication network is overloaded in the first time window, or refer to the steps involved in step 230 to calculate whether the communication network is overloaded in other time windows). In this way, it is possible to continue to detect whether there will still be warning information in the communication network adjusted by the response strategy, so as to achieve continuous monitoring and adjustment of the communication network until no warning information is generated. Thus, after step 2602, the data processing method may also include steps 2603 to 2605.

[0082] Step 2603: Based on the response strategy, adjust the communication network and determine whether the adjusted communication network generates target early warning information during the third time period.

[0083] Step 2604: If the target early warning information is determined to be generated, the network management system generates target influencing factor information of the target network area on the service according to the service type of the service carried by the target network area corresponding to the target early warning information; and regenerates the response strategy to adjust the communication network according to the target influencing factor information, so as to continuously adjust the communication network.

[0084] Step 2605: If it is determined that no target warning information has been generated, process the service traffic data through the adjusted communication network.

[0085] In some embodiments of this application, a network early warning distribution visualization map can be generated based on early warning information and a network capacity distribution visualization map, so that users can view overloaded network areas in the communication network. Based on this, after step 2502 or step 230, the data processing method may further include steps 2701 and 2702.

[0086] Step 2701: Map the early warning information onto the network capacity distribution visualization map to obtain the network early warning distribution visualization map. The network early warning distribution visualization map is used to represent the overloaded network areas and the degree of overload in the network capacity distribution visualization map during the first time period.

[0087] Step 2702: Upon receiving the warning cancellation information corresponding to the warning information, update the network warning distribution visualization map to obtain the updated network warning distribution visualization map. The updated network warning distribution visualization map does not display the overloaded network areas and the degree of network area overload information in the network capacity distribution visualization map during the first time period.

[0088] Therefore, this application embodiment can establish a communication network model that comprehensively describes network capacity. Employing machine learning methods, it achieves adaptive modeling of time-varying network traffic characteristics, enabling long-term prediction and evaluation of communication network capacity, as well as accurate network capacity prediction and early warning of potential network faults. This results in intelligent and precise monitoring and management of communication network capacity. Thus, by combining machine learning technology with the field of communication networks, it is expected to solve the problem of poor adaptability to dynamic network changes in current methods, achieving proactive monitoring and fault early warning of network capacity.

[0089] To better illustrate the data processing method provided in the embodiments of this application, based on Figure 1 and Figure 2 The content shown is specifically combined with Figure 3 Please provide a detailed explanation.

[0090] (1) Construct a communication network topology model: collect network connection information, establish a digital network model, and obtain key network parameters.

[0091] Specifically, 1) Collect network connection information: Obtain data on the connection relationships between nodes in the network, such as the link bandwidth between nodes, connection matrix, etc.

[0092] 2) Constructing a network graph: Based on the connection relationships, a digital model of the network is constructed using graph theory to represent the network topology.

[0093] 3) Identify nodes: Identify key nodes in the network diagram, such as regional nodes and core landing points, and obtain information such as their type and location.

[0094] 4) Identify links: Identify the connection links between nodes and obtain parameters such as link length, bandwidth, and delay.

[0095] 5) Store the network model: Store the constructed network topology diagram as data for easy access later.

[0096] 6) Update the model: When the network topology changes, update the network model in a timely manner to ensure that it reflects the current network status.

[0097] Therefore, by collecting network connection information, constructing a digital network model using graph theory, storing and continuously updating it, we can achieve digital modeling of the communication network topology and provide basic network data for subsequent capacity forecasting.

[0098] (2) Monitor node status: Real-time detection of static attributes and dynamic status parameters of each node.

[0099] Specifically, 1) Identify core monitoring nodes: Based on the network topology, identify key regional nodes, core nodes, and access devices as monitoring nodes. These nodes directly affect the distribution of network capacity.

[0100] 2) Determine monitoring indicators: Monitor the static attributes and dynamic status parameters of the nodes, including static indicators such as processing capacity, number of ports, and channel configuration, as well as dynamic indicators such as port traffic, CPU and memory usage, and service routing.

[0101] 3) Deploy monitoring agents: On the selected core monitoring nodes, deploy lightweight monitoring agents to collect node status data. The monitoring agents periodically obtain node parameters and report them to the data acquisition module.

[0102] 4) Construct a node status dataset: After aggregating the status data of all monitored nodes in the data acquisition module and removing redundant and erroneous data, a node status dataset is formed for training machine learning models.

[0103] 5) Data preprocessing: Perform preprocessing such as normalization on the node state data to eliminate the impact of data bias on model training.

[0104] 6) Incremental training of the model: Using newly collected data, the machine learning model is incrementally trained to continuously adapt to changes in the network state.

[0105] 7) Model retraining: When there is a major upgrade to the network topology or node hardware, it is necessary to recollect large-scale state data and retrain the model.

[0106] Therefore, by deploying monitoring agents at key nodes to collect node status parameters in real time and processing the data, data reflecting the current network status can be obtained, providing a basis for capacity forecasting.

[0107] (3) Collect network traffic data: Collect historical and real-time business traffic data in the network.

[0108] Specifically, 1) Determine monitoring points: Determine monitoring points at network nodes and important service entry points to obtain representative network traffic data.

[0109] 2) Deploy traffic detectors: Deploy hardware and software traffic detection systems at monitoring points to capture flowing data packets and parse packet headers to obtain traffic parameters.

[0110] 3) Analyze the service type: Determine the service type by analyzing the packet header information, such as video traffic, file download, web page access, etc.

[0111] 4) Traffic parameters statistics: Parse traffic parameters such as source address, destination address, packet size, and arrival time, and perform statistical analysis.

[0112] 5) Construct traffic samples: Extract a large number of business flow statistical parameters to form a sample dataset that reflects the characteristics of network traffic.

[0113] 6) Perform data augmentation: Use methods such as sliding time windows and mixed flow to augment the original samples and increase the number of samples.

[0114] 7) Extracting time-series features: Analyze traffic flow time-series patterns and extract dynamic time-series features such as periodicity and suddenness as input to the model.

[0115] 8) Divide the dataset into training and testing sets: Divide the traffic sample dataset into training dataset and testing dataset for training and testing machine learning models.

[0116] 9) Continuously collect new samples: Monitor newly emerging traffic on the network, collect new samples for incremental training of the model, and update it as the network changes.

[0117] Therefore, by measuring and statistically analyzing traffic, sample data containing dynamic characteristics of network traffic can be obtained, providing a large amount of effective training data for building learning models.

[0118] (4) Establish a machine learning prediction model: Use a recurrent neural network to train and obtain a model of the mapping relationship between network traffic and capacity status.

[0119] Specifically, 1) Determine the model algorithm: For network traffic time series data, select a recurrent neural network algorithm that can handle time-related data, such as LSTM, GRU, etc.

[0120] 2) Constructing the network structure: Design the neural network structure with input layer, multiple hidden layers and output layer, and determine the number of nodes in each layer.

[0121] 3) Determine the training data: Integrate the network topology model, node status data, and traffic data into samples to serve as model training data.

[0122] 4) Set model parameters: Determine the parameters of the network layers, such as the weight matrix and threshold function parameters, and set the training hyperparameters, such as the number of iterations and the batch size.

[0123] 5) Model training: Using the training data, train the network model through the backpropagation algorithm, minimize the loss function, and obtain the model parameters.

[0124] 6) Model evaluation: Use independent test data to evaluate the model's prediction accuracy, recall, and other metrics.

[0125] 7) Model optimization: Optimize model performance by adjusting network structure, training parameters, etc.

[0126] 8) Model Deployment: Implement model services in the network, receive real-time network status data, and perform traffic-to-capacity mapping prediction.

[0127] 9) Model update: Through incremental training, the model is continuously optimized using newly collected data to adapt to changes in the network.

[0128] 10) Model Management: Manage different versions of models, use the optimal model or model combination as needed, and achieve a controllable model lifecycle.

[0129] Therefore, by using machine learning modeling, the relationship between traffic and capacity status in the network can be learned, enabling accurate capacity prediction based on current network conditions.

[0130] (5) Input network state parameters: Input the network state parameters detected in real time into the machine learning model.

[0131] (6) Conduct network capacity forecasting: Based on the current network status, use machine learning models to predict the capacity load of each node and link in the network.

[0132] Specifically, 1) Receive topology model: Obtain the current network topology and parameter data from the topology modeling module.

[0133] 2) Receive node status: Obtain real-time status data of each node from the node monitoring module, such as port traffic and service routing.

[0134] 3) Receive service traffic: Obtain real-time service traffic data entering the network from the traffic monitoring module.

[0135] 4) Data preprocessing: Cleaning, anomaly removal, and normalization are performed on the received data.

[0136] 5) Constructing input vectors: Organize the processed data into the input vector format defined by the machine learning model.

[0137] 6) Model initialization: Load the pre-trained network capacity prediction model and initialize the model's weights and parameters.

[0138] 7) Perform model prediction: Input the constructed input vector into the initialized model, run the model's forward computation process, and obtain the output vector.

[0139] 8) Parse capacity output: Parse the output vector to extract the predicted network capacity information, including the expected load status of each node and link.

[0140] 9) Generate a capacity map: Map the predicted capacity information onto the network topology to form an intuitive visualization map of the network capacity distribution.

[0141] 10) Push capacity results: Push the analyzed and visualized capacity prediction results to the subsequent network analysis module.

[0142] Therefore, by comprehensively processing real-time network status data, accurate network capacity prediction based on intelligent models can be achieved.

[0143] (7) Determine if the capacity is overloaded: Determine if the predicted capacity is close to the threshold.

[0144] Specifically, 1) Set thresholds: Based on network capacity and historical operating data, set capacity thresholds for nodes and links. Exceeding the threshold is considered overload.

[0145] 2) Comparative capacity forecasts: The network capacity forecasts predicted by the machine learning model are compared with the set thresholds item by item.

[0146] 3) Identify overloaded areas: In the comparison results, find the network nodes or links whose predicted capacity exceeds the threshold, and determine the specific network areas that are overloaded.

[0147] 4) Assess overload level: Calculate the difference between predicted capacity and threshold to assess the severity of network overload.

[0148] 5) Comprehensive overload factors: Consider the importance of the overloaded parts, the type of real-time service, and other factors to conduct a comprehensive analysis of the impact of overload.

[0149] 6) Generate early warning information: Based on the overload location, severity, and influencing factors, generate clear network overload early warning information.

[0150] 7) Push early warning notifications: Proactively push early warning information to the network management system in the form of messages.

[0151] 8) Early warning display: In the network monitoring system, the network area experiencing overload is displayed in a visual way.

[0152] 9) Continuous tracking and early warning: When the overload is relieved, a notification to relieve the early warning is sent to realize continuous tracking of the network overload status.

[0153] 10) Save early warning records: Keep records of network overload predictions and early warnings for system optimization.

[0154] By comparing predictions with thresholds, the system can anticipate and warn of potential network overload, thus aiding in network management.

[0155] (8) Sending early warning information: If the capacity is overloaded, send a capacity early warning information to the network management.

[0156] (9) The network management responds to the warning by taking countermeasures such as expanding network capacity or scheduling business traffic.

[0157] Specifically, 1) Receiving early warning information: The network management system receives network overload early warnings pushed by the early warning module.

[0158] 2) Locate the overloaded area: View the location of the overloaded node or link indicated in the warning in the network topology system.

[0159] 3) Analyze the impact on services: Based on the types of services carried by the overloaded links, analyze the degree of impact of the overload on services.

[0160] 4) Develop response strategies: Based on the cause, impact, and urgency of the overload, develop corresponding strategies such as capacity expansion, migration, or service restriction.

[0161] 5) Increase network resources: If the overload is due to insufficient node or link capacity, increase network resources such as bandwidth and transmission equipment.

[0162] 6) Service traffic rerouting: If overload occurs due to routing problems, some service traffic can be rerouted to other links.

[0163] 7) Adjust service priorities: Reduce the network priority of non-critical services and reserve broadband resources for important services.

[0164] 8) Limit traffic for a specific service: If necessary, you can choose to limit or isolate the traffic of services that have a significant impact on overloaded nodes.

[0165] 9) Update load balancing strategy: Re-formulate business configuration strategy to guide traffic around overloaded nodes.

[0166] 10) Issue control commands: Issue the control commands defined in the policy to network devices to perform traffic management.

[0167] Therefore, by taking targeted measures in response to early warning information, network overload can be effectively eliminated or mitigated, ensuring stable business operations.

[0168] (10) Continuously monitor the network to achieve closed-loop control.

[0169] Specifically, 1) Maintain network status monitoring: While dealing with network overload, continuously monitor network status through node agents and traffic detectors.

[0170] 2) Obtain control command response: Obtain control commands such as resource expansion and traffic scheduling issued by the network management system, and execution feedback.

[0171] 3) Verify the effect of capacity improvement: After the network status changes, re-predict the capacity to verify whether the network capacity has been improved.

[0172] 4) Determine overload status: If the capacity is still overloaded after the first round of warnings, a second or even multiple rounds of warnings and controls are needed.

[0173] 5) End the warning: If the capacity forecast indicates that the overload has been relieved, stop the warning and report the problem to the management for resolution.

[0174] 6) Save control cases: Save the management's control instructions and effects as cases for future reference in handling problems.

[0175] 7) Analyze the root cause: Combine network traffic and topology change information to analyze the root cause of network overload.

[0176] 8) Propose planning recommendations: If the overload is caused by insufficient network capacity, provide medium- to long-term planning recommendations for network expansion or adjustment.

[0177] 9) Optimize early warning rules: Continuously optimize early warning rules by analyzing historical cases to make them more intelligent and efficient.

[0178] 10) Improve user experience: Establishing a closed-loop system for status monitoring, predictive early warning, and network control can significantly improve the user's business experience.

[0179] By constructing a closed-loop mechanism, the network can be continuously monitored, and multiple rounds of early warning and control can be carried out based on the prediction results until the network overload problem is resolved. At the same time, the root cause analysis of the problem can be completed to carry out in-depth network optimization.

[0180] Therefore, through this process, network status can be continuously monitored, network capacity can be predicted in real time, early warnings can be issued before potential overload, network management can be guided, and network stability and controllability can be achieved.

[0181] The data processing method provided in this application employs machine learning to adaptively model and learn the complex dynamic characteristics of the network, offering advantages over fixed models. It not only achieves short-term prediction but also assesses long-term network capacity trends to guide network planning. Furthermore, it enables proactive monitoring and accurate prediction of network capacity, rather than passively waiting for problems to occur. When potential faults are predicted, early warnings and responses can be provided to ensure network stability and reliability. A closed-loop adaptive network control framework is established, offering superior overall performance compared to traditional methods. It comprehensively considers multiple factors, such as topology, traffic, node status, and service impact, resulting in a better user experience and effectively preventing the impact of network faults on services. It is also more versatile and can be applied to various types of communication networks, demonstrating innovation. No solution for monitoring communication network capacity using machine learning has been seen in this field before. Thus, while achieving proactive and accurate network fault prediction, it also improves network stability.

[0182] This application also provides a data processing apparatus, specifically combined with... Figure 4 Please provide a detailed explanation.

[0183] Figure 4 This is a schematic diagram of the structure of a data processing apparatus provided in one embodiment of this application.

[0184] In some embodiments of this application, Figure 4 The data processing device shown can be installed in the computing device provided in the embodiments of this application.

[0185] like Figure 4 As shown, the data processing device 40 may specifically include:

[0186] The acquisition module 401 is used to acquire network information of the communication network. The communication network has a network topology structure, which includes N nodes and a first link connecting every two nodes. The network information includes the structural parameters of the network topology structure, the node status information of the N nodes, the service traffic data entering the communication network, and the network status parameters. The network status parameters include the parameters generated during the process of the communication network processing the service traffic data.

[0187] Calculation module 402 is used to calculate the first network capacity information of the communication network in the first time period based on network information and through a network capacity prediction model;

[0188] The mapping module 403 is used to map the first network capacity information to the network topology to obtain a network capacity distribution visualization map. The network capacity distribution visualization map includes P nodes out of N nodes and a second link for connecting every two nodes out of P nodes. The network capacity distribution visualization map is used to characterize the first load state of each node out of P nodes in the first time period and the second load state of the second link in the first time period, where P∈[1,N].

[0189] Therefore, in this embodiment, the data processing device can predict the network capacity information of the communication network based on various factors such as the structural parameters of the network topology, the node status information of the nodes, the service traffic data entering the communication network, and network status parameters, through a network capacity prediction model. This is beneficial for accurately generating a dynamic mapping relationship between the service traffic data processed by the communication network in the first time period and the load status of each node in the communication network in the first time period. This can improve the adaptability to dynamic and complex network traffic changes, thereby improving the accuracy of calculating the long-term communication network capacity under dynamic and complex network conditions. This achieves the purpose of proactively and accurately monitoring the network capacity information of the communication network and discovering potential faults. Furthermore, by mapping the first network capacity information to the network topology, the network capacity distribution visualization map can reflect the load status of each node in the first time period and the load status of the link connecting at least two nodes in the first time period, allowing users to intuitively and clearly view the service traffic data processed by each node and link in the first time period.

[0190] The data processing device 40 in the embodiments of this application will be described in detail below.

[0191] In some embodiments of this application, the data processing device 40 may further include a first generation module; wherein,

[0192] The calculation module 402 can also be used to calculate the network capacity threshold of the communication network based on the network capacity of the communication network and the second network capacity information of the communication network in the second time period, wherein the end time of the second time period is earlier than the start time of the first time period.

[0193] The first generation module is used to generate early warning information when the first network capacity value corresponding to the first network capacity information is greater than or equal to the network capacity threshold. The early warning information is used to characterize the overload of the communication network in the first time period.

[0194] In some embodiments of this application, the calculation module 402 may be specifically used to calculate the degree of overload of the communication network in the first time period based on the difference between the first network capacity value and the network capacity threshold when the first network capacity value corresponding to the first network capacity information is greater than or equal to the network capacity threshold.

[0195] Based on the degree information and the network area of ​​the communication network that is overloaded in the first time period, an early warning information is generated. The network area includes at least one of the following: a first area where the nodes in the communication network are located, and a second area where the links connecting every two nodes in the communication network are located.

[0196] In some embodiments of this application, the first generation module may be specifically used to determine the first region where the node in the communication network is located as the network region when the first network capacity value includes the first sub-network capacity value and the second sub-network capacity value, the network capacity threshold includes the first network capacity threshold and the second network capacity threshold, and the first sub-network capacity value is greater than or equal to the first network capacity threshold.

[0197] Specifically, the first generation module can be used to determine the second region where the link connecting every two nodes in the communication network is located as the network region when the first network capacity value includes the first sub-network capacity value and the second sub-network capacity value, the network capacity threshold includes the first network capacity threshold and the second network capacity threshold, and the second sub-network capacity value is greater than or equal to the second network capacity threshold.

[0198] Specifically, the first generation module can be used to determine the first region and the second region as network regions when the first network capacity value includes the first sub-network capacity value and the second sub-network capacity value, the network capacity threshold includes the first network capacity threshold and the second network capacity threshold, and the first sub-network capacity value is greater than or equal to the first network capacity threshold and the second sub-network capacity value is greater than or equal to the second network capacity threshold.

[0199] In some embodiments of this application, the data processing device 40 may further include an update module; wherein,

[0200] The mapping module 403 can also be used to map the early warning information onto the network capacity distribution visualization map to obtain the network early warning distribution visualization map. The network early warning distribution visualization map is used to represent the overloaded network areas and the degree of overload in the network capacity distribution visualization map in the first time period.

[0201] The update module is used to update the network early warning distribution visualization map when the early warning cancellation information corresponding to the early warning information is received, so as to obtain the updated network early warning distribution visualization map. The updated network early warning distribution visualization map does not display the overloaded network areas and the degree of overload of the network areas in the network capacity distribution visualization map in the first time period.

[0202] In some embodiments of this application, the data processing device 40 may further include a second generation module; wherein,

[0203] The second generation module is used to generate information on the factors affecting services in a network area based on the service type of the services carried by the network area through the network management system.

[0204] Based on information on influencing factors, generate response strategies.

[0205] In some embodiments of this application, the influencing factors include at least one of the following: the number of nodes in the communication network is lower than a first preset threshold, the link capacity in the communication network is lower than a second preset threshold, the service routing failure in the network area, the bandwidth resources of the communication network are lower than a third preset threshold, and the service configuration strategy of the communication network is abnormal.

[0206] Based on this, the second generation module can specifically be used to generate a response strategy for increasing nodes in the communication network when the influencing factor information is that the number of nodes in the communication network is lower than a first preset threshold.

[0207] When the influencing factor information is that the link capacity in the communication network is lower than the second preset threshold, a response strategy for increasing the link bandwidth in the communication network is generated.

[0208] In the event of a service routing failure in a network area, a response strategy is generated to redirect service traffic data processed in the network area to a target network area in the communication network. The target network area does not overlap with the network area.

[0209] When the influencing factor information is that the bandwidth resources of the communication network are lower than the third preset threshold, a response strategy is generated to adjust the network priority for the communication network to process service traffic data.

[0210] When the influencing factor information is that the service configuration policy of the communication network is abnormal, generate response strategies for restricting or isolating the service traffic data processed by nodes in the network area, generate response strategies for updating the load balance of the communication network, or generate response strategies for issuing instructions to manage the processing of service traffic data of the communication network.

[0211] In some embodiments of this application, the data processing device 40 may further include an adjustment module, a third generation module, and a processing module; wherein,

[0212] The adjustment module is used to adjust the communication network based on the response strategy and determine whether the adjusted communication network generates target early warning information in the third time period.

[0213] The third generation module is used to generate target influencing factor information of the target network area on the service according to the service type of the target network area corresponding to the target warning information, through the network management system; and to regenerate the response strategy to adjust the communication network based on the target influencing factor information, so as to continuously adjust the communication network.

[0214] The processing module is used to process service traffic data through the adjusted communication network when it is determined that no target early warning information has been generated.

[0215] In some embodiments of this application, the data processing device 40 may further include a filtering module; wherein,

[0216] A filtering module is used to filter network capacity prediction models from at least two candidate network capacity prediction models, based on the communication network, when the network capacity prediction model includes at least two candidate network capacity prediction models. The number of network capacity prediction models is at least one.

[0217] In some embodiments of this application, the structural parameters include at least one of the following: the node type of each of the N nodes, and the node identifier of the first link connection node;

[0218] The node status information includes at least one of the following: port traffic of each of the N nodes, and service routing information of each of the N nodes;

[0219] Network status parameters include at least one of the following: service latency for processing service traffic data, and service packet loss rate for processing service traffic data.

[0220] In some embodiments of this application, the data processing device 40 may further include a training module; wherein,

[0221] The acquisition module 401 can also be used to acquire sample information, which includes sample network information of each of the M sample communication networks. The sample communication network has a sample network topology, which includes K sample nodes and a first sample link connecting every two sample nodes in the K sample nodes. The sample network information includes sample structure parameters of the sample network topology, sample node information of the K sample nodes, sample service traffic data entering the sample communication network, and sample network status parameters. The sample network status parameters include parameters generated during the processing of sample service traffic data by the sample communication network. M is an integer greater than 1, and K∈[1,M].

[0222] The training module is used to train the network capacity prediction model based on the sample information until the preset training conditions are met, thus obtaining the network capacity prediction model.

[0223] In some embodiments of this application, the training module can also be used to train the network capacity prediction model based on the updated sample information when the sample information is updated, until the preset training conditions are met, and obtain an optimized network capacity prediction model. The optimized network capacity prediction model is used to calculate the network capacity information of the communication network in the fourth time period.

[0224] In some embodiments of this application, the acquisition module 401 may be specifically used to mark the sample nodes and the sample links used to connect each pair of sample nodes in the sample network topology map according to the connection information of each sample node in each sample communication network, so as to obtain the sample structure parameters of the sample network topology.

[0225] Select K sample nodes from the sample network topology and obtain the node information of each of the K sample nodes; determine the node information of each of the K sample nodes as the sample node information of the K sample nodes;

[0226] Traffic parameters are obtained through a traffic detector, and the sample service type is determined based on the traffic parameters; sample service traffic data is generated based on the sample service information and traffic parameters related to the sample service type.

[0227] Monitor the parameters generated during the processing of sample service traffic data by the sample communication network, and determine the parameters generated during the processing of sample service traffic data by the sample communication network as sample network status parameters.

[0228] Based on the same inventive concept, this application also provides a computer device. (Specifically combined with...) Figure 5 Please provide a detailed explanation.

[0229] Figure 5 This is a schematic diagram of the structure of a computer device provided in one embodiment of this application.

[0230] like Figure 5 As shown, the computer device may include at least one of the following as described in the embodiments of this application: an electronic device, a server. The computer device may include a processor 501 and a memory 502 storing computer program instructions.

[0231] Specifically, the processor 501 may include a central processing unit (CPU), or an application-specific integrated circuit (ASTC), or one or more integrated circuits that can be configured to implement the embodiments of this application.

[0232] Memory 502 may include mass storage for data or instructions. For example, and not limitingly, memory 502 may include a hard disk drive (HDD), floppy disk drive, flash memory, optical disk drive, magneto-optical disk drive, magnetic tape drive, or Universal Serial Bus (USB) drive, or a combination of two or more of these. Where appropriate, memory 502 may include removable or non-removable (or fixed) media. Where appropriate, memory 502 may be internal or external to a computer device. In a particular embodiment, memory 502 is non-volatile solid-state memory. In a particular embodiment, memory 502 includes solid-state storage (ROM). Where appropriate, the ROM may be a mask-programmed ROM, a programmable ROM (PROM), an erasable PROM (EPROM), an electrically erasable PROM (EEPROM), an electrically rewritable ROM (EAROM), or flash memory, or a combination of two or more of these.

[0233] The processor 501 implements any of the data processing methods described in the above embodiments by reading and executing computer program instructions stored in the memory 502.

[0234] In one example, the computer device may also include a communication interface 503 and a bus 510. Wherein, as... Figure 5 As shown, the processor 501, memory 502, and communication interface 503 are connected through bus 510 and complete communication with each other.

[0235] The communication interface 503 is mainly used to realize communication between various modules, devices, units and / or equipment in the embodiments of this application.

[0236] Bus 510 includes hardware, software, or both, that couples components of a flow control device together. For example, and not limitingly, the bus may include an Accelerated Graphics Port (AGP) or other graphics bus, an Enhanced Industry Standard System (ETSA) bus, a Front Side Bus (FSB), an HyperTransport (HT) interconnect, an Industry Standard System (TSA) bus, an Infinite Bandwidth Interconnect, a Low Pin Count (LPC) bus, a memory bus, a Microchannel System (MCA) bus, a Peripheral Component Interconnect (PCT) bus, a PCT-Express (PCT-X) bus, a Serial Advanced Technology Attachment (SATA) bus, a Video Electronics Standards Association Local (VLB) bus, or other suitable buses, or combinations of two or more of these. Where appropriate, bus 510 may include one or more buses. Although specific buses are described and illustrated in embodiments of this application, this application contemplates any suitable bus or interconnect.

[0237] The data processing device can execute the data processing method described in the embodiments of this application, thereby achieving the combination Figures 1 to 5 The data processing methods and apparatus described.

[0238] Furthermore, in conjunction with the data processing methods in the above embodiments, this application embodiment can provide a computer-readable storage medium for implementation. This computer-readable storage medium stores computer program instructions; when executed by a processor, these computer program instructions implement any of the data processing methods in the above embodiments.

[0239] It should be clarified that this application is not limited to the specific configurations and processes described above and shown in the figures. For the sake of brevity, detailed descriptions of known methods are omitted here. In the above embodiments, several specific steps are described and shown as examples. However, the method process of this application is not limited to the specific steps described and shown. Those skilled in the art can make various changes, modifications, and additions, or change the order of steps, after understanding the spirit of this application.

[0240] The functional blocks shown in the above block diagram can be implemented as hardware, software, firmware, or a combination thereof. When implemented in hardware, they can be, for example, electronic circuits, application-specific integrated circuits (ASTCs), appropriate firmware, plug-ins, function cards, etc. When implemented in software, the elements of this application are programs or code segments used to perform the required tasks. Programs or code segments can be stored on a machine-readable medium or transmitted over a transmission medium or communication link via data signals carried on a carrier wave. "Machine-readable medium" can include any medium capable of storing or transmitting information. Examples of machine-readable media include electronic circuits, semiconductor memory devices, ROM, flash memory, erasable ROM (EROM), floppy disks, CD-ROMs, optical disks, hard disks, fiber optic media, radio frequency (RF) links, etc. Code segments can be downloaded via computer networks such as the Internet, intranets, etc.

[0241] It should also be noted that the exemplary embodiments mentioned in this application describe methods or systems based on a series of steps or apparatus. However, this application is not limited to the order of the above steps; that is, the steps can be performed in the order mentioned in the embodiments, or in a different order, or several steps can be performed simultaneously.

[0242] The above are merely specific embodiments of this application. Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, modules, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here. It should be understood that the protection scope of this application is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope disclosed in this application, and these modifications or substitutions should all be covered within the protection scope of this application.

Claims

1. A data processing method, characterized in that, include: The network information of the communication network is obtained. The communication network has a network topology structure, which includes N nodes and a first link for connecting every two nodes among the N nodes. The network information includes the structural parameters of the network topology structure, the node status information of the N nodes, the service traffic data entering the communication network, and the network status parameters. The network status parameters include parameters generated by the communication network during the processing of the service traffic data. Based on the network information, the first network capacity information of the communication network in the first time period is calculated using the network capacity prediction model; The first network capacity information is mapped onto the network topology to obtain a network capacity distribution visualization map. The network capacity distribution visualization map includes P nodes out of the N nodes and a second link for connecting every two nodes out of the P nodes. The network capacity distribution visualization map is used to characterize the first load state of each of the P nodes in the first time period and the second load state of the second link in the first time period, where P∈[1,N]. Before calculating the first network capacity information of the communication network based on the network information and using a network capacity prediction model, the method further includes: Obtain sample information, which includes sample network information for each of the M sample communication networks. The sample communication network has a sample network topology, which includes K sample nodes and a first sample link connecting every two sample nodes among the K sample nodes. The sample network information includes sample structure parameters of the sample network topology, sample node information of the K sample nodes, sample service traffic data entering the sample communication network, and sample network status parameters. The sample network status parameters include parameters generated by the sample communication network during the processing of the sample service traffic data. M is an integer greater than 1, and K∈[1,M]. Based on the sample information, the sample network capacity prediction model is trained until the preset training conditions are met, thus obtaining the network capacity prediction model.

2. The method according to claim 1, characterized in that, The method further includes: Based on the network capacity of the communication network and the second network capacity information of the communication network in the second time period, the network capacity threshold of the communication network is calculated, wherein the end time of the second time period is earlier than the start time of the first time period; If the first network capacity value corresponding to the first network capacity information is greater than or equal to the network capacity threshold, an early warning message is generated. The early warning message is used to characterize that the communication network is overloaded during the first time period.

3. The method according to claim 2, characterized in that, The step of generating early warning information when the first network capacity value corresponding to the first network capacity information is greater than or equal to the network capacity threshold includes: If the first network capacity value corresponding to the first network capacity information is greater than or equal to the network capacity threshold, the degree of overload of the communication network during the first time period is calculated based on the difference between the first network capacity value and the network capacity threshold. The warning information is generated based on the degree information and the network area of ​​the communication network that is overloaded during the first time period. The network area includes at least one of the following: a first area where the nodes in the communication network are located, and a second area where the link connecting every two nodes in the communication network is located.

4. The method according to claim 3, characterized in that, The first network capacity value includes a first sub-network capacity value and a second sub-network capacity value, and the network capacity threshold includes a first network capacity threshold and a second network capacity threshold; Before generating the warning information based on the degree information and the overloaded network area of ​​the communication network during the first time period, the method further includes: If the capacity value of the first sub-network is greater than or equal to the capacity threshold of the first network, the first region where the node in the communication network is located is determined as the network region. If the capacity value of the second sub-network is greater than or equal to the capacity threshold of the second network, the second region where the link connecting every two nodes in the communication network is located is determined as the network region. If the capacity value of the first sub-network is greater than or equal to the first network capacity threshold, and the capacity value of the second sub-network is greater than or equal to the second network capacity threshold, then the first region and the second region are determined as the network region.

5. The method according to any one of claims 2-4, characterized in that, The method further includes: The early warning information is mapped onto the network capacity distribution visualization map to obtain a network early warning distribution visualization map. The network early warning distribution visualization map is used to represent the overloaded network areas and the degree of overload in the network capacity distribution visualization map during the first time period. Upon receiving warning cancellation information corresponding to the warning information, the network warning distribution visualization map is updated to obtain an updated network warning distribution visualization map. The updated network warning distribution visualization map does not display the overloaded network areas and the degree of overload of the network areas in the network capacity distribution visualization map during the first time period.

6. The method according to claim 3, characterized in that, The method further includes: The network management system generates information on the factors influencing the network area on the services based on the service types carried by the network area. Based on the information on the influencing factors, a response strategy is generated.

7. The method according to claim 6, characterized in that, The influencing factors include at least one of the following: the number of nodes in the communication network is lower than a first preset threshold, the link capacity in the communication network is lower than a second preset threshold, the service routing in the network area is faulty, the bandwidth resources of the communication network are lower than a third preset threshold, and the service configuration strategy of the communication network is abnormal. The step of generating a response strategy based on the influencing factor information includes: If the number of nodes in the communication network is lower than a first preset threshold, a response strategy for increasing the number of nodes in the communication network is generated. When the influencing factor information indicates that the link capacity in the communication network is lower than a second preset threshold, a response strategy for increasing the link bandwidth in the communication network is generated. In the event that the influencing factor information indicates a service routing failure in the network area, a response strategy is generated to redirect the service traffic data processed in the network area to a target network area of ​​the communication network, wherein the target network area does not overlap with the network area. When the influencing factor information indicates that the bandwidth resources of the communication network are lower than a third preset threshold, a response strategy is generated to adjust the network priority for the communication network in processing service traffic data. When the influencing factor information indicates that the service configuration policy of the communication network is abnormal, a response strategy is generated to restrict or isolate the service traffic data processed by nodes in the network area, a response strategy is generated to update the load balancing of the communication network, or a response strategy is generated to issue instructions to manage the processing of the service traffic data by the communication network.

8. The method according to claim 7, characterized in that, The method further includes: Based on the aforementioned response strategy, the communication network is adjusted, and it is determined whether the adjusted communication network generates target early warning information during the third time period. If the target early warning information is determined to be generated, the network management system generates target influencing factor information of the target network area on the service according to the service type of the service carried by the target network area corresponding to the target early warning information; and regenerates the response strategy to adjust the communication network according to the target influencing factor information, so as to continuously adjust the communication network. If it is determined that the target warning information has not been generated, the service traffic data is processed through the adjusted communication network.

9. The method according to claim 1, characterized in that, The network capacity prediction model includes at least two candidate network capacity prediction models; Before calculating the first network capacity information of the communication network based on the network information and using the network capacity prediction model, the method further includes: Based on the communication network, a network capacity prediction model is selected from at least two candidate network capacity prediction models, wherein the number of network capacity prediction models is at least one.

10. The method according to claim 1, characterized in that, The structural parameters include at least one of the following: the node type of each of the N nodes, and the node identifier of the first link connection node; The node status information includes at least one of the following: port traffic of each of the N nodes, and service routing information of each of the N nodes; The network status parameters include at least one of the following: service latency for processing the service traffic data, and service packet loss rate for processing the service traffic data.

11. The method according to claim 1, characterized in that, The method further includes: When the sample information is updated, the network capacity prediction model is trained based on the updated sample information until the preset training conditions are met, and an optimized network capacity prediction model is obtained. The optimized network capacity prediction model is used to calculate the network capacity information of the communication network in the fourth time period.

12. The method according to claim 1, characterized in that, The acquisition of sample information includes: Based on the connection information of each sample node in each sample communication network, the sample nodes and the sample links used to connect every two sample nodes are marked in the sample network topology diagram to obtain the sample structure parameters of the sample network topology. K sample nodes are selected from the sample network topology, and the node information of each of the K sample nodes is obtained; the node information of each of the K sample nodes is determined as the sample node information of the K sample nodes. Traffic parameters are obtained through a traffic detector, and the sample service type is determined based on the traffic parameters; the sample service traffic data is generated based on the sample service information related to the sample service type and the traffic parameters. The parameters generated during the processing of the sample service traffic data by the sample communication network are monitored, and the parameters generated during the processing of the sample service traffic data by the sample communication network are determined as the sample network status parameters.

13. A data processing apparatus, characterized in that, include: The acquisition module is used to acquire network information of a communication network. The communication network has a network topology structure, which includes N nodes and a first link for connecting every two nodes among the N nodes. The network information includes structural parameters of the network topology structure, node status information of the N nodes, service traffic data entering the communication network, and network status parameters. The network status parameters include parameters generated by the communication network during the processing of the service traffic data. The calculation module is used to calculate the first network capacity information of the communication network in the first time period based on the network information and through the network capacity prediction model; The mapping module is used to map the first network capacity information to the network topology to obtain a network capacity distribution visualization map. The network capacity distribution visualization map includes P nodes out of the N nodes and a second link for connecting every two nodes out of the P nodes. The network capacity distribution visualization map is used to characterize the first load state of each of the P nodes in the first time period and the second load state of the second link in the first time period, where P∈[1,N]. Before performing the step of calculating the first network capacity information of the communication network based on the network information and through the network capacity prediction model, the acquisition module is further configured to acquire sample information, which includes sample network information of each of the M sample communication networks, the sample communication network having a sample network topology, the sample network topology including K sample nodes and a first sample link connecting every two sample nodes in the K sample nodes, the sample network information including sample structure parameters of the sample network topology, sample node information of the K sample nodes, sample service traffic data entering the sample communication network and sample network status parameters, the sample network status parameters including parameters generated during the processing of the sample service traffic data by the sample communication network, where M is an integer greater than 1 and K∈[1,M]; The training module is used to train the sample network capacity prediction model based on the sample information until the preset training conditions are met, thereby obtaining the network capacity prediction model.

14. A computer device, characterized in that, The device includes: a processor and a memory storing computer program instructions; When the processor executes the computer program instructions, it implements the steps of the data processing method as described in any one of claims 1-12.

15. A readable storage medium, characterized in that, The readable storage medium stores a program or instructions that, when executed by a processor, implement the steps of the data processing method as described in any one of claims 1-12.

16. A computer program product, characterized in that, The program product is stored in a storage medium, and the program product is executed by at least one processor to implement the steps of the data processing method as described in any one of claims 1-12.

Images