Network traffic bandwidth real-time dynamic adjustment method and related equipment thereof

Abstract

The embodiment of the invention belongs to the technical field of artificial intelligence, and relates to a network traffic bandwidth real-time dynamic adjustment method and related equipment thereof. Training an AI intelligent model used for real-time prediction of network traffic; transmitting the real-time flow at each network node to the A I intelligent model; obtaining network prediction traffic; and carrying out network flow bandwidth adjustment on each network node in combination with the network flow bandwidth adjustment strategy. AI intelligent model training is carried out by adopting an ant colony algorithm, hyper-parameter combination selection is carried out, and high availability and prediction accuracy of the trained A I intelligent model are ensured. The method is applied to business fields such as financial insurance and medical online services needing network flow regulation and control, offline service scheduling can be carried out in combination with specific offline and online businesses, meanwhile, the network flow can be reasonably distributed, and the situation that an online service system crashes due to unreasonable network flow distribution is avoided.

Description

Technical Field

[0001] This application relates to the field of artificial intelligence technology, and in particular, to a method for real-time dynamic adjustment of network traffic bandwidth and related devices. Background Art

[0002] Currently, in business scenarios such as online services, since more and more processing needs to combine online traffic for node allocation processing. For example, in the above-mentioned property insurance application service system such as online auto insurance, it is necessary to first declare an insurance application, allocate seat assignment docking personnel nodes, declare people submit proof document nodes, review nodes, etc. In the business scenario of auto insurance purchase or renewal, most car owners' car purchase times are concentrated around March every year. Therefore, there are certain off-peak and peak seasons for auto insurance renewal. Obviously, there are significant differences in the network traffic (tasks) borne by the renewal nodes during peak and off-peak seasons; another example is the appointment queuing nodes, doctor assignment nodes, payment processing nodes, medicine self-pickup window assignment nodes, etc. in the medical service system. Obviously, the appointment or processing traffic of different nodes is different at different time points. How to dynamically allocate network traffic for different processing nodes has become a major problem.

[0003] Currently, there are already optimization methods based on network traffic prediction. For example, for mobile user services, operators classify users into priority users and non-priority users. To ensure better service quality for priority users, some solutions use machine learning methods to predict user traffic patterns, and use two supervised learning methods, namely the Naive Bayes classifier and logistic regression, to predict the possible time when priority users are in an active state, and apply the prediction results to the user scheduling strategy for effective bandwidth management. However, the optimization of this prediction process is to use supervised learning methods to make decision control on the admission of service requests, without using the dynamic perception of the neural network for service characteristics, and without considering the long-term and short-term comprehensive changes of network traffic, resulting in inaccurate network traffic predicted by the existing network traffic prediction AI intelligent model at the target network node. Summary of the Invention

[0004] The purpose of the embodiments of this application is to propose a method for real-time dynamic adjustment of network traffic bandwidth and related devices to solve the problem that the network traffic predicted by the existing network traffic prediction AI intelligent model at the target network node is not accurate enough.

[0005] In a first aspect, the embodiments of this application provide a method for real-time dynamic adjustment of network traffic bandwidth, adopting the following technical solutions:

[0006] A method for real-time dynamic adjustment of network traffic bandwidth includes the following steps:

[0007] Monitor and obtain the real-time traffic characterization data at each network node in the target system during the target time period;

[0008] Perform amplification processing on the real-time traffic characterization data at each of the network nodes using a unified amplification method to obtain a batch of sample data;

[0009] Randomly divide the batch of sample data according to a preset ratio value to obtain a training set and a validation set;

[0010] Input the training set and the validation set into a long short-term memory neural network based on the ant colony algorithm to train an AI intelligent model for real-time network traffic prediction;

[0011] Establish a data transmission relationship between the trained AI intelligent model for real-time network traffic prediction and the real-time traffic monitoring components preset for each of the network nodes;

[0012] Based on the data transmission relationship, transmit the real-time traffic at each of the network nodes recorded by the real-time traffic monitoring components to the AI intelligent model for real-time network traffic prediction;

[0013] Obtain the network prediction traffic corresponding to each of the network nodes at a future target time point output by the AI intelligent model;

[0014] According to the network prediction traffic corresponding to each of the network nodes at a future target time point and a preset network traffic bandwidth adjustment strategy, adjust the network traffic bandwidth of each of the network nodes at the future target time point.

[0015] In a second aspect, an embodiment of the present application further provides a device for real-time dynamic adjustment of network traffic bandwidth, which adopts the following technical solutions:

[0016] A device for real-time dynamic adjustment of network traffic bandwidth includes:

[0017] A real-time traffic characterization data acquisition module, configured to monitor and acquire the real-time traffic characterization data at each of the network nodes in a target system during a target time period;

[0018] A real-time traffic characterization data amplification module, configured to perform amplification processing on the real-time traffic characterization data at each of the network nodes using a unified amplification method to obtain a batch of sample data;

[0019] A training set and validation set division module, configured to randomly divide the batch of sample data according to a preset ratio value to obtain a training set and a validation set;

[0020] An AI intelligent model training module, configured to input the training set and the validation set into a long short-term memory neural network based on the ant colony algorithm to train an AI intelligent model for real-time network traffic prediction;

[0021] A data transmission relationship establishment module, configured to establish a data transmission relationship between the trained AI intelligent model for real-time network traffic prediction and the real-time traffic monitoring components respectively preset for each network node;

[0022] A real-time traffic record result transmission module, configured to transmit the real-time traffic at each network node recorded by the real-time traffic monitoring component to the AI intelligent model for real-time network traffic prediction based on the data transmission relationship;

[0023] A network prediction traffic output module, configured to obtain the network prediction traffic corresponding to each network node at a future target time point output by the AI intelligent model;

[0024] A network traffic bandwidth adjustment module, configured to adjust the network traffic bandwidth of each network node at the future target time point according to the network prediction traffic corresponding to each network node at the future target time point and a preset network traffic bandwidth adjustment strategy.

[0025] In a third aspect, an embodiment of the present application further provides a computer device, which adopts the following technical solution:

[0026] A computer device includes a memory and a processor. Computer-readable instructions are stored in the memory, and when the processor executes the computer-readable instructions, the steps of the above-mentioned network traffic bandwidth real-time dynamic adjustment method are implemented.

[0027] In a fourth aspect, an embodiment of the present application further provides a computer-readable storage medium, which adopts the following technical solution:

[0028] A computer-readable storage medium stores computer-readable instructions, and when the computer-readable instructions are executed by a processor, the steps of the network traffic bandwidth real-time dynamic adjustment method as described above are implemented.

[0029] Compared with the prior art, the embodiments of the present application mainly have the following beneficial effects:

[0030] The real-time dynamic adjustment method for network traffic bandwidth in the embodiments of this application monitors and obtains the real-time traffic characterization data at each network node in the target system within the target time period; performs amplification processing to obtain a batch of sample data; randomly divides it to obtain a training set and a validation set; inputs them into a long short-term memory neural network based on the ant colony algorithm to train an AI intelligent model for real-time network traffic prediction; transmits the real-time traffic at each network node recorded by the real-time traffic monitoring component to the AI intelligent model for real-time network traffic prediction; obtains the network predicted traffic corresponding to each network node at the future target time point; and combines the network traffic bandwidth adjustment strategy to adjust the network traffic bandwidth of each network node at the future target time point. Using the ant colony algorithm for AI intelligent model training ensures the high availability and prediction accuracy of the trained AI intelligent model. Applying the network traffic bandwidth dynamic adjustment method to the financial insurance business scenario or the online medical service business can reasonably perform offline service scheduling and allocate the network traffic of each processing node in combination with the situations of different business nodes, avoid the problem of unreasonable network traffic allocation by the business processing system, and improve the stability of the business system. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] To more clearly illustrate the solutions in this application, the following will briefly introduce the drawings required for the description of the embodiments of this application. Obviously, the following-described drawings are some embodiments of this application. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0032] Figure 1 is an exemplary system architecture diagram to which this application can be applied;

[0033] Figure 2 is a flowchart of an embodiment of the real-time dynamic adjustment method for network traffic bandwidth according to this application;

[0034] Figure 3 is Figure 2 a flowchart of a specific embodiment of step 201;

[0035] Figure 4 is a flowchart of a specific embodiment of updating the recording result in the real-time dynamic adjustment method for network traffic bandwidth described in this application;

[0036] Figure 5 is Figure 2 a flowchart of a specific embodiment of step 202;

[0037] Figure 6 is Figure 2 a flowchart of a specific embodiment of step 204;

[0038] Figure 7 is Figure 6 A flowchart of a specific embodiment of step 604;

[0039] Figure 8 is Figure 7 A flowchart of a specific embodiment of step 705;

[0040] Figure 9 A flowchart of a specific embodiment for correcting the real-time traffic at each network node recorded by the real-time traffic monitoring component in the method for real-time dynamic adjustment of network traffic bandwidth described in this application;

[0041] Figure 10 is Figure 2 A flowchart of a specific embodiment of step 208;

[0042] Figure 11 A schematic structural diagram of an embodiment of the network traffic bandwidth real-time dynamic adjustment device according to this application;

[0043] Figure 12 A schematic structural diagram of an embodiment of the computer device according to this application. Detailed implementation manners

[0044] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those of ordinary skill in the technical field to which this application belongs; the terms used in the specification of this application are only for the purpose of describing specific embodiments and are not intended to limit this application; the terms "including" and "having" and any variations thereof in the specification and claims of this application and the above drawings are intended to cover non-exclusive inclusion. The terms "first", "second", etc. in the specification and claims of this application or the above drawings are used to distinguish different objects and not to describe a specific order.

[0045] Reference to "embodiment" herein means that a particular feature, structure, or characteristic described in connection with the embodiment can be included in at least one embodiment of this application. The phrase does not necessarily refer to the same embodiment at every occurrence in the specification, nor is it an independent or alternative embodiment mutually exclusive with other embodiments. Those skilled in the art will explicitly and implicitly understand that the embodiments described herein can be combined with other embodiments.

[0046] To enable those skilled in the technical field to better understand the solution of this application, the technical solutions in the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings.

[0047] Such as Figure 1As shown, the system architecture 100 may include a terminal device 101, a network 102, and a server 103. The terminal device 101 may be a laptop computer 1011, a tablet computer 1012, or a mobile phone 1013. The network 102 is a medium for providing a communication link between the terminal device 101 and the server 103. The network 102 may include various connection types, such as wired, wireless communication links, or fiber optic cables, etc.

[0048] Users can use the terminal device 101 to interact with the server 103 through the network 102 to receive or send messages, etc. Various communication client applications may be installed on the terminal device 101, such as a web browser application, a shopping application, a search application, an instant messaging tool, an email client, a social platform software, etc.

[0049] The terminal device 101 may be various electronic devices with a display screen and supporting web browsing. In addition to the laptop computer 1011, the tablet computer 1012, or the mobile phone 1013, the terminal device 101 may also be an e-book reader, an MP3 player (Moving Picture Experts Group Audio Layer III), an MP4 (Moving Picture Experts Group Audio Layer IV) player, a laptop portable computer, a desktop computer, etc.

[0050] The server 103 may be a server providing various services, such as a background server supporting the pages displayed on the terminal device 101.

[0051] It should be noted that the network traffic bandwidth real-time dynamic adjustment method provided by the embodiments of the present application is generally executed by the server. Correspondingly, the network traffic bandwidth real-time dynamic adjustment device is generally set in the server.

[0052] It should be understood that Figure 1 the numbers of terminal devices, networks, and servers in

[0053] Continue to refer to Figure 2 , which shows a flowchart of an embodiment of the network traffic bandwidth real-time dynamic adjustment method according to the present application. The network traffic bandwidth real-time dynamic adjustment method includes the following steps:

[0054] Step 201, monitor and obtain the real-time traffic characterization data at each network node in the target system within the target time period.

[0055] In this embodiment, the target time period includes a preset one year, half year, etc. The target system is, for example, a financial insurance business system, and each network node is, for example, a vehicle insurance promotion node, a vehicle insurance application node, a vehicle insurance renewal node, an insurance claim settlement node, etc. in the above financial insurance business system. Applying the network traffic bandwidth dynamic adjustment method to the financial insurance business scenario can combine the off-peak and peak seasons of different business nodes, so as to reasonably allocate the network traffic of each processing node, avoid the problem of unreasonable network traffic allocation in the financial business processing system, and increase the stability of the business system.

[0056] In this embodiment, the target time period further includes within a preset one hour, one day, half day, etc. The specific start and end time points are freely set according to the actual business needs; the target system is, for example, an online medical service system, and each network node is, for example, an appointment queuing node, a doctor assignment node, a payment processing node, a drug self-service window assignment node, etc. in the above online medical service system.

[0057] By monitoring and obtaining the real-time traffic characterization data at each network node in the target system within the target time period, applying the network traffic bandwidth real-time dynamic adjustment method to the field of online medical services can combine the specific offline and online medical services, predict which medical processing node is the busiest at which time point, and accordingly perform offline service scheduling. At the same time, it can also predict in which time periods medical patients are concentrated in online service operations, so as to reasonably allocate the network traffic of each processing node and avoid the collapse of the online medical service system due to unreasonable network traffic allocation.

[0058] Step 202: Perform amplification processing on the real-time traffic characterization data at each network node by using a unified amplification method to obtain a batch of sample data.

[0059] By performing amplification processing on the real-time traffic characterization data at each network node by using a unified amplification method, it not only ensures the consistency of the amplification method but also ensures the amplification from a small amount of monitored data to a batch of sample data, so as to ensure the sufficiency of the training data when performing model training later, and thus train a more accurate prediction model.

[0060] Step 203: Randomly divide the batch of sample data according to a preset ratio value to obtain a training set and a validation set.

[0061] In this embodiment, the preset ratio value is, for example, 9:1 or 8:2, and the random division can be specifically implemented by using a random sampling method. For example, first randomly select 9 sample data from the batch of sample data and put them into the training set, and then randomly select 1 sample data and put it into the validation set; repeat this process until the set training sample size is reached and the random division is completed.

[0062] Step 204: Input the training set and the validation set into the long short-term memory neural network based on the ant colony algorithm to train an AI intelligent model for real-time network traffic prediction.

[0063] In this embodiment, the long short-term memory neural network is, for example, an LSTM neural network; the long short-term memory neural network based on the ant colony algorithm means that when training the long short-term memory neural network, the optimal solution screening method of the ant colony algorithm is used to screen out the optimal model hyperparameter combination from the target number of model hyperparameter combinations, and the trained long short-term memory neural network corresponding to the optimal model hyperparameter combination is used as the AI intelligent model for real-time network traffic prediction.

[0064] Step 205: Establish a data transmission relationship between the trained AI intelligent model for real-time network traffic prediction and the real-time traffic monitoring components preset for each network node.

[0065] By establishing a data transmission relationship between the trained AI intelligent model for real-time network traffic prediction and the real-time traffic monitoring components preset for each network node, when actually using the AI intelligent model, the real-time traffic at each network node recorded by the real-time traffic monitoring components can be directly transmitted to the AI intelligent model for real-time network traffic prediction.

[0066] Step 206: Based on the data transmission relationship, transmit the real-time traffic at each network node recorded by the real-time traffic monitoring components to the AI intelligent model for real-time network traffic prediction.

[0067] Step 207: Obtain the network prediction traffic corresponding to each network node at the future target time point output by the AI intelligent model.

[0068] Step 208: According to the network prediction traffic corresponding to each network node at the future target time point and the preset network traffic bandwidth adjustment strategy, adjust the network traffic bandwidth of each network node at the future target time point.

[0069] By combining the network traffic prediction results at specific time points, the network traffic bandwidth of each network node is adjusted, avoiding the collapse of the target business system. Applying the real-time dynamic adjustment method of the network traffic bandwidth to the field of online medical services can combine specific offline and online medical services to predict which medical processing node is the busiest at which time point, and accordingly perform offline service scheduling. At the same time, it can also predict the time periods when medical patients are concentrated in online service operations, so as to reasonably allocate the network traffic of each processing node and avoid the collapse of the online medical service system due to unreasonable network traffic allocation. Of course, the real-time dynamic adjustment method of the network traffic bandwidth can also be applied to the field of financial business to predict that the online operations of financial stocks and investment users are relatively frequent at a certain time point, so as to reasonably adjust the network traffic bandwidth.

[0070] In this embodiment, by monitoring and obtaining the real-time traffic characterization data at each network node in the target system during the target time period; performing amplification processing to obtain a batch of sample data; randomly dividing to obtain a training set and a validation set; inputting them into a long short-term memory neural network based on the ant colony algorithm to train an AI intelligent model for real-time network traffic prediction; transmitting the real-time traffic at each network node recorded by the real-time traffic monitoring component to the AI intelligent model for real-time network traffic prediction; obtaining the network prediction traffic corresponding to each network node at the future target time point; and combining the network traffic bandwidth adjustment strategy to adjust the network traffic bandwidth of each network node at the future target time point. The ant colony algorithm is used for training the AI intelligent model, and hyperparameter combinations are selected to ensure the high availability and prediction accuracy of the trained AI intelligent model.

[0071] Continue to refer to Figure 3 , Figure 3 is Figure 2 The flowchart of a specific embodiment of step 201 includes:

[0072] Step 301, start the real-time traffic monitoring components preset for each network node respectively;

[0073] Step 302, according to the real-time traffic monitoring component and the node number information of each network node, record the real-time traffic corresponding to each network node at each unit time point, where each unit time point includes each second;

[0074] Step 303, based on the recording results, draw the real-time traffic change curves corresponding to each network node;

[0075] Step 304, set the real-time traffic change curves corresponding to each network node as the real-time traffic characterization data at each network node.

[0076] By converting discrete continuous data, i.e., the real-time traffic record results, into real-time traffic change curves, it is convenient for subsequent processing in combination with line chart methods, and it is more reasonable to learn the changes of network traffic at each network node from the long-term and short-term dependencies and the changes of the entire business system.

[0077] Continue to refer to Figure 4 , in some alternative implementation manners, before performing the step 303, the method further includes a step of updating the record results. Figure 4 FIG. is a flowchart of a specific embodiment of updating the record results in the real-time dynamic adjustment method of network traffic bandwidth described in the present application, including the following steps:

[0078] Step 401, perform outlier detection on the record results according to a preset outlier detection method to obtain an outlier detection result, where the outliers include real-time traffic values with obvious recording errors and obvious missing recordings;

[0079] Step 402, combine the real-time traffic corresponding to the previous unit time point and the next unit time point of the target outlier, and perform correction processing on each outlier in the outlier detection result;

[0080] Step 403, supplement the correction processing result to the record results to update the record results.

[0081] By correcting and updating the record results, the scientificity and accuracy of the real-time traffic change curves corresponding to each network node drawn subsequently are ensured.

[0082] Continue to refer to Figure 5 , Figure 5 is Figure 2 a flowchart of a specific embodiment of the step 202, including:

[0083] Step 501, input the real-time traffic change curves corresponding to each network node into a preset sampling processing component;

[0084] Step 502, perform sliding sampling on the real-time traffic change curves corresponding to each network node according to the preset window size and step length in the sampling processing component to obtain the real-time traffic change curves corresponding to each network node in different time periods;

[0085] Specifically, the window size refers to the horizontal length value of the real-time traffic change curve intercepted each time. For example, the real-time traffic change curve with a duration of half an hour is intercepted each time. The step size refers to the interval between the starting positions when intercepting continuously twice. For example, when intercepting for the first time, the start time is 12:10, and the window length is 20 minutes. Then the real-time traffic change curve intercepted for the first time is the real-time traffic change curve corresponding to the period from 12:10 to 12:30. Assuming the step size is 5 minutes, when intercepting for the second time, the start time is 12:15. Then the real-time traffic change curve intercepted for the second time is the real-time traffic change curve corresponding to the 20 minutes from 12:15 to 12:35.

[0086] Step 503: Obtain the real-time traffic change curves corresponding to each network node in different time periods as the batch sample data after amplification processing, where the start and end times of the real-time traffic change curves corresponding to all network nodes in each sample data are the same.

[0087] By performing amplification processing on the real-time traffic characterization data at each network node, it is ensured that the small amount of monitoring data is amplified to a batch of sample data, so as to ensure the sufficiency of the training data during subsequent model training, and thus train a more accurate prediction model.

[0088] Continue to refer to Figure 6 , Figure 6 is Figure 2 The flowchart of a specific embodiment of step 204 includes:

[0089] Step 601: Identify the model hyperparameters to be optimized in the long short-term memory neural network, where the model hyperparameters to be optimized include the number of hidden units, learning rate, and number of training epochs.

[0090] Step 602: Use the random combination method and the preset optional values of the model hyperparameters to generate a target number of hyperparameter combinations, where the preset optional values of the model hyperparameters include the optional values corresponding to the number of hidden units, learning rate, and number of training epochs respectively.

[0091] Step 603: Set the target number of hyperparameter combinations as the ant colony in the ant algorithm.

[0092] Step 604: Sequentially select the hyperparameter combinations corresponding to different ant individuals in the ant colony as the training parameters of the long short-term memory neural network, and perform network training in combination with the training set and the validation set.

[0093] Step 605: Based on the training results, select the hyperparameter combination corresponding to the minimum loss value of the verification results corresponding to the validation set as the target hyperparameter combination, where the minimum value is less than a preset loss threshold;

[0094] Step 606: Use the long short-term memory neural network trained based on the target hyperparameter combination as the AI intelligent model for real-time prediction of network traffic.

[0095] Adopt the optimal solution screening method of the ant colony algorithm to screen out the optimal model hyperparameter combination from the target number of model hyperparameter combinations, and use the trained long short-term memory neural network corresponding to the optimal model hyperparameter combination as the AI intelligent model for real-time prediction of network traffic to ensure the high availability and prediction accuracy of the trained AI intelligent model.

[0096] Continue to refer to Figure 7 , Figure 7 Yes Figure 6 The flowchart of a specific embodiment of step 604 includes:

[0097] Step 701: Set the hyperparameter combination corresponding to the current ant individual as the training parameters of the long short-term memory neural network to complete the initial deployment of the long short-term memory neural network;

[0098] Step 702: Input the training set into the long short-term memory neural network with the initial deployment completed to perform learning and training on the long short-term memory neural network;

[0099] Step 703: Stop the learning and training of the long short-term memory neural network until the learning rate corresponding to the long short-term memory neural network reaches the learning rate in the hyperparameter combination and the number of training cycles also reaches the number of training cycles in the hyperparameter combination, and obtain the long short-term memory neural network to be verified;

[0100] Step 704: Input the validation set into the long short-term memory neural network to be verified, and obtain the real-time traffic prediction change curves at each network node in the target system output by the long short-term memory neural network to be verified;

[0101] Step 705: Based on the real-time traffic prediction change curves at each network node in the target system and the real-time traffic actual change curves at each network node in the validation set, obtain the loss value corresponding to the hyperparameter combination;

[0102] Step 706: Successively use different ant individuals as the current ant individual, and repeat steps 701 to 705 until the loss values corresponding to all hyperparameter combinations are obtained to complete the network training.

[0103] Continue to refer to Figure 8 , Figure 8 is Figure 7 a flowchart of a specific embodiment of step 705, including:

[0104] Step 801: Identify the time period for the actual change curve of the real-time traffic at each network node in the validation set, and obtain the time period corresponding to each real-time traffic change curve;

[0105] Step 802: According to the time period corresponding to each real-time traffic change curve and the network node corresponding to each real-time traffic change curve, obtain the real-time traffic prediction change curve corresponding to the network node in the same time period from the corresponding real-time traffic prediction change curve;

[0106] Step 803: Superimpose and compare the real-time traffic prediction change curve corresponding to the same network node in the same time period with the actual change curve of the real-time traffic;

[0107] Step 804: For the non-superimposed curve part, mark it with a different color, and count the total proportion value of the time of the curve part finally marked with the different color;

[0108] Step 805: Use the total proportion value of the superimposed time as the loss value corresponding to the hyperparameter combination.

[0109] By adopting the curve superposition method and the different color marking method to calculate the loss value corresponding to the hyperparameter combination, it is more intuitive than the previous method of calculating the loss value in the form of numerical results, and there is no need to discretize and serialize a large amount of data, saving machine processing resources.

[0110] Continue to refer to Figure 9 , in some alternative implementation manners, before executing step 206, the method further includes a step of correcting the real-time traffic at each network node recorded by the real-time traffic monitoring component, Figure 9 is a flowchart of a specific embodiment of correcting the real-time traffic at each network node recorded by the real-time traffic monitoring component in the real-time network traffic bandwidth dynamic adjustment method of the present application, including the following steps:

[0111] Step 901: According to a preset outlier detection method, perform outlier detection on the real-time traffic at each network node recorded by the real-time traffic monitoring component to obtain an outlier detection result, where the outlier includes an obvious recording error and an obvious missed recording real-time traffic value;

[0112] Step 902: Combine the real-time traffic corresponding to the previous and next unit time points of the target outlier, and perform correction processing on each outlier in the outlier detection result;

[0113] Step 903: Supplement the correction processing result into the real-time traffic at each network node recorded by the real-time traffic monitoring component to update the recorded result.

[0114] By correcting and updating the real-time traffic at each network node recorded by the real-time traffic monitoring component, the accuracy of the network prediction traffic corresponding to each network node obtained during subsequent network traffic prediction is ensured.

[0115] Continue to refer to Figure 10 , Figure 10 Yes Figure 2 The flowchart of a specific embodiment of step 208 includes:

[0116] Step 10a: Identify the network traffic bandwidth pre-allocated for each network node respectively as the allocated bandwidth;

[0117] Step 10b: If the network prediction traffic corresponding to the current network node at the future target time point exceeds the actual traffic allowed to pass through the allocated bandwidth corresponding to the current network node, perform dynamic expansion processing on the allocated bandwidth at the future target time point;

[0118] Step 10c: If the network prediction traffic corresponding to the current network node at the future target time point does not exceed the actual traffic allowed to pass through the allocated bandwidth corresponding to the current network node, and the network prediction traffic exceeds half of the allowed actual traffic, the allocated bandwidth remains unchanged at the future target time point;

[0119] Step 10d: If the network prediction traffic corresponding to the current network node at the future target time point does not exceed the actual traffic allowed to pass through the allocated bandwidth corresponding to the current network node, and the network prediction traffic does not exceed half of the allowed actual traffic, perform dynamic capacity reduction processing on the allocated bandwidth at the future target time point.

[0120] In this embodiment, the step of dynamically expanding the allocated bandwidth at the future target time point specifically includes: according to the proportional relationship between the network predicted traffic and the actual traffic allowed to pass, adopting a progressive multiple expansion method for the allocated bandwidth to perform expansion processing until the actual traffic allowed to pass by the latest bandwidth after expansion is greater than the network predicted traffic, and then stop the expansion processing. Wherein, each expansion processing will make the actual traffic allowed to pass by the latest bandwidth after expansion be 2 times that before expansion.

[0121] In this embodiment, the step of dynamically reducing the allocated bandwidth at the future target time point specifically includes: according to the proportional relationship between the network predicted traffic and the actual traffic allowed to pass, adopting a progressive multiple reduction method for the allocated bandwidth to perform reduction processing until the actual traffic allowed to pass by the latest bandwidth after reduction is greater than the network predicted traffic and less than 2 times the network predicted traffic, and then stop the reduction processing. Wherein, each reduction processing will make the actual traffic allowed to pass by the latest bandwidth after reduction be 0.5 times that before reduction.

[0122] By combining the network traffic bandwidths respectively allocated to each network node in advance, and the network predicted traffic corresponding to each network node at the future target time point, it is convenient to dynamically adjust the network traffic bandwidths of each network node in the target system during the network traffic transmission process, and ensure the stability of the system.

[0123] This application monitors and obtains the real-time traffic characterization data at each network node in the target system within the target time period; performs amplification processing to obtain a batch of sample data; randomly divides to obtain a training set and a validation set; inputs them into a long short-term memory neural network based on the ant colony algorithm to train an AI intelligent model for real-time network traffic prediction; transmits the real-time traffic at each network node recorded by the real-time traffic monitoring component to the AI intelligent model for real-time network traffic prediction; obtains the network predicted traffic corresponding to each network node at the future target time point respectively; combines the network traffic bandwidth adjustment strategy to adjust the network traffic bandwidths of each network node at the future target time point. The ant colony algorithm is used for training the AI intelligent model and performing hyperparameter combination selection to ensure the high availability and prediction accuracy of the trained AI intelligent model.

[0124] The embodiments of this application can acquire and process relevant data based on artificial intelligence technology. Among them, artificial intelligence (AI) is a theory, method, technology and application system that uses a digital computer or a machine controlled by a digital computer to simulate, extend and expand human intelligence, perceive the environment, acquire knowledge and use knowledge to obtain the best results.

[0125] The basic technologies of artificial intelligence generally include technologies such as sensors, dedicated artificial intelligence chips, cloud computing, distributed storage, big data processing technologies, operation / interaction systems, and mechatronics. The software technologies of artificial intelligence mainly include several major directions such as computer vision technology, robotics, biometric technology, speech processing technology, natural language processing technology, and machine learning / deep learning.

[0126] In the embodiments of the present application, by monitoring and obtaining the real-time traffic characterization data at each network node in the target system within the target time period; performing amplification processing to obtain a batch of sample data; randomly dividing to obtain a training set and a validation set; inputting into a long short-term memory neural network based on the ant colony algorithm to train an AI intelligent model for real-time network traffic prediction; transmitting the real-time traffic at each network node recorded by the real-time traffic monitoring component to the AI intelligent model for real-time network traffic prediction; obtaining the network predicted traffic corresponding to each network node at the future target time point; and combining the network traffic bandwidth adjustment strategy to adjust the network traffic bandwidth of each network node at the future target time point. The ant colony algorithm is used for training the AI intelligent model, and hyperparameter combination selection is performed to ensure the high availability and prediction accuracy of the trained AI intelligent model.

[0127] Further referring to Figure 11 and as an implementation of the method shown above Figure 2 , the present application provides an embodiment of a device for real-time dynamic adjustment of network traffic bandwidth. This device embodiment corresponds to the method embodiment shown in Figure 2 and this device can be specifically applied to various electronic devices.

[0128] As shown in Figure 11 , the device 11 for real-time dynamic adjustment of network traffic bandwidth described in this embodiment includes: a real-time traffic characterization data acquisition module 11a, a real-time traffic characterization data amplification module 11b, a training set and validation set division module 11c, an AI intelligent model training module 11d, a data transmission relationship establishment module 11e, a real-time traffic record result transmission module 11f, a network predicted traffic output module 11g, and a network traffic bandwidth adjustment module 11h. Among them:

[0129] The real-time traffic characterization data acquisition module 11a is used to monitor and obtain the real-time traffic characterization data at each network node in the target system within the target time period;

[0130] The real-time traffic characterization data amplification module 11b is used to perform amplification processing on the real-time traffic characterization data at each network node by using a unified amplification method to obtain a batch of sample data;

[0131] The training set and validation set division module 11c is used to randomly divide the batch sample data according to a preset ratio value to obtain a training set and a validation set;

[0132] The AI intelligent model training module 11d is used to input the training set and the validation set into a long short-term memory neural network based on the ant colony algorithm to train an AI intelligent model for real-time network traffic prediction;

[0133] The data transmission relationship establishment module 11e is used to establish a data transmission relationship between the trained AI intelligent model for real-time network traffic prediction and the real-time traffic monitoring components preset for each network node;

[0134] The real-time traffic record result transmission module 11f is used to transmit the real-time traffic at each network node recorded by the real-time traffic monitoring component to the AI intelligent model for real-time network traffic prediction based on the data transmission relationship;

[0135] The network prediction traffic output module 11g is used to obtain the network prediction traffic corresponding to each network node at a future target time point output by the AI intelligent model;

[0136] The network traffic bandwidth adjustment module 11h is used to adjust the network traffic bandwidth of each network node at the future target time point according to the network prediction traffic corresponding to each network node at the future target time point and a preset network traffic bandwidth adjustment strategy.

[0137] In this application, the real-time traffic representation data at each network node in the target system is monitored and obtained during the target time period; amplified processing is performed to obtain batch sample data; randomly divided to obtain a training set and a validation set; input into a long short-term memory neural network based on the ant colony algorithm to train an AI intelligent model for real-time network traffic prediction; the real-time traffic at each network node recorded by the real-time traffic monitoring component is transmitted to the AI intelligent model for real-time network traffic prediction; the network prediction traffic corresponding to each network node at a future target time point is obtained; combined with the network traffic bandwidth adjustment strategy, the network traffic bandwidth of each network node is adjusted at the future target time point. The ant colony algorithm is used for AI intelligent model training and hyperparameter combination selection to ensure the high availability and prediction accuracy of the trained AI intelligent model.

[0138] In this embodiment, the network traffic bandwidth real-time dynamic adjustment device 11 further includes: an outlier detection module, an outlier correction module, and a network traffic record result update module. Among them:

[0139] An outlier detection module, configured to perform outlier detection on the recording result according to a preset outlier detection method, so as to obtain an outlier detection result, where the outliers include obvious recording errors and real-time traffic values with obvious missing records;

[0140] An outlier correction module, configured to correct each outlier in the outlier detection result by combining the real-time traffic corresponding to the previous unit time point and the next unit time point of the target outlier;

[0141] A network traffic recording result update module, configured to supplement the correction processing result into the recording result to update the recording result.

[0142] In this embodiment, the real-time traffic characterization data amplification module 11b includes a sampling input unit, a sliding sampling processing unit, and a batch sample data acquisition unit. Among them:

[0143] The sampling input unit is configured to input the real-time traffic change curves corresponding to the respective network nodes into a preset sampling processing component;

[0144] The sliding sampling processing unit is configured to perform sliding sampling on the real-time traffic change curves corresponding to the respective network nodes according to the window size and step length preset in the sampling processing component, so as to obtain the real-time traffic change curves corresponding to the respective network nodes in different time periods;

[0145] The batch sample data acquisition unit is configured to obtain the real-time traffic change curves corresponding to the respective network nodes in different time periods as the amplified batch sample data, where the start and end times of the real-time traffic change curves corresponding to all network nodes in each sample data are the same.

[0146] In this embodiment, the AI intelligent model training module 11d further includes: a time period identification and acquisition unit, a corresponding predicted change curve acquisition unit, a superposition and comparison unit, a marking and calculation unit, and a loss value determination unit. Among them:

[0147] The time period identification and acquisition unit is configured to perform time period identification on the actual real-time traffic change curves at each network node in the validation set, so as to obtain the time period corresponding to each real-time traffic change curve;

[0148] The corresponding predicted change curve acquisition unit is configured to obtain the real-time traffic predicted change curve corresponding to the corresponding network node in the same time period from the corresponding real-time traffic predicted change curves according to the time period corresponding to each real-time traffic change curve and the network node corresponding to each real-time traffic change curve;

[0149] A superposition and comparison unit, configured to superpose and compare the real-time traffic prediction change curve and the real-time traffic actual change curve corresponding to the same network node in the same time period;

[0150] A marking and calculating unit, configured to mark the non-superposed curve parts with different colors and calculate the total proportion value of the time of the curve parts finally marked with different colors in the total superposition time;

[0151] A loss value determination unit, configured to use the total superposition time proportion value as the loss value corresponding to the hyperparameter combination.

[0152] In this embodiment, the network traffic bandwidth adjustment module 11h further includes a dynamic capacity expansion processing unit, configured to perform capacity expansion processing on the allocated bandwidth in a progressive multiple expansion manner according to the proportional relationship between the network predicted traffic and the actual traffic allowed to pass, until the actual traffic allowed to pass by the latest bandwidth after expansion is greater than the network predicted traffic, and then stop the capacity expansion processing, where each capacity expansion processing makes the actual traffic allowed to pass by the latest bandwidth after expansion twice that before expansion.

[0153] In this embodiment, the network traffic bandwidth adjustment module 11h further includes a dynamic capacity reduction processing unit, configured to perform capacity reduction processing on the allocated bandwidth in a progressive multiple reduction manner according to the proportional relationship between the network predicted traffic and the actual traffic allowed to pass, until the actual traffic allowed to pass by the latest bandwidth after reduction is greater than the network predicted traffic and less than twice the network predicted traffic, and then stop the capacity reduction processing, where each capacity reduction processing makes the actual traffic allowed to pass by the latest bandwidth after reduction 0.5 times that before reduction.

[0154] Those of ordinary skill in the art can understand that all or part of the processes in the methods of the above embodiments can be completed by instructing relevant hardware through computer-readable instructions, and the computer-readable instructions can be stored in a computer-readable storage medium. When the program is executed, it can include the processes of the embodiments of the above methods. Among them, the foregoing storage medium can be a non-volatile storage medium such as a magnetic disk, an optical disk, a read-only memory (ROM), or a random access memory (RAM), etc.

[0155] It should be understood that although the steps in the flowchart of the accompanying drawings are shown sequentially in the direction of the arrows, these steps are not necessarily executed sequentially in the order indicated by the arrows. Unless otherwise clearly stated in this document, there is no strict order restriction for the execution of these steps, and they can be executed in other orders. Moreover, at least some of the steps in the flowchart of the accompanying drawings may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily executed at the same time, but can be executed at different times, and their execution order is not necessarily sequential, but can be executed alternately or in turn with at least some of the sub-steps or stages of other steps or other steps.

[0156] To solve the above technical problems, an embodiment of the present application also provides a computer device. For details, please refer to Figure 12 , Figure 12 which is the basic structural block diagram of the computer device in this embodiment.

[0157] The computer device 12 includes a memory 12a, a processor 12b, and a network interface 12c that are communicatively connected to each other through a system bus. It should be noted that Figure 12 only the computer device 12 with components such as a memory 12a, a processor 12b, and a network interface 12c is shown, but it should be understood that it is not required to implement all the shown components, and more or fewer components can be alternatively implemented. Among them, those skilled in the art of this technology can understand that a computer device here is a device that can automatically perform numerical calculations and / or information processing according to pre-set or stored instructions, and its hardware includes but is not limited to microprocessors, application specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), digital signal processors (DSPs), embedded devices, etc.

[0158] The computer device can be a computing device such as a desktop computer, a notebook, a palm computer, and a cloud server. The computer device can perform human-computer interaction with users through means such as a keyboard, a mouse, a remote control, a touchpad, or a voice control device.

[0159] The memory 12a at least includes one type of readable storage medium, which includes flash memory, hard disk, multimedia card, card-type memory (such as SD or DX memory, etc.), random access memory (RAM), static random access memory (SRAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), programmable read-only memory (PROM), magnetic memory, magnetic disk, optical disk, etc. In some embodiments, the memory 12a may be an internal storage unit of the computer device 12, such as the hard disk or memory of the computer device 12. In other embodiments, the memory 12a may also be an external storage device of the computer device 12, such as a plug-in hard disk equipped on the computer device 12, a Smart Media Card (SMC), a Secure Digital (SD) card, a Flash Card, etc. Of course, the memory 12a may also include both the internal storage unit and the external storage device of the computer device 12. In this embodiment, the memory 12a is generally used to store the operating system and various application software installed on the computer device 12, such as computer-readable instructions for a method of real-time dynamic adjustment of network traffic bandwidth. In addition, the memory 12a can also be used to temporarily store various data that have been output or will be output.

[0160] In some embodiments, the processor 12b may be a central processing unit (CPU), a controller, a microcontroller, a microprocessor, or other data processing chips. The processor 12b is generally used to control the overall operation of the computer device 12. In this embodiment, the processor 12b is used to run the computer-readable instructions stored in the memory 12a or process data, such as running the computer-readable instructions for the method of real-time dynamic adjustment of network traffic bandwidth.

[0161] The network interface 12c may include a wireless network interface or a wired network interface, and this network interface 12c is generally used to establish a communication connection between the computer device 12 and other electronic devices.

[0162] The computer device proposed in this embodiment belongs to the field of artificial intelligence technology and is applied to the network traffic regulation scenario in online services such as medical health or fintech. This application monitors and obtains the real-time traffic characterization data at each network node in the target system within the target time period; performs amplification processing to obtain a batch of sample data; randomly divides it to obtain a training set and a validation set; inputs them into a long short-term memory neural network based on the ant colony algorithm to train an AI intelligent model for real-time network traffic prediction; transmits the real-time traffic at each network node recorded by the real-time traffic monitoring component to the AI intelligent model for real-time network traffic prediction; obtains the network predicted traffic corresponding to each network node at the future target time point; and combines the network traffic bandwidth adjustment strategy to adjust the network traffic bandwidth of each network node at the future target time point. The ant colony algorithm is used for AI intelligent model training, and hyperparameter combination selection is performed to ensure the high availability and prediction accuracy of the trained AI intelligent model.

[0163] This application also provides another implementation manner, that is, to provide a computer-readable storage medium storing computer-readable instructions that can be executed by a processor to cause the processor to execute the steps of the above real-time dynamic adjustment method for network traffic bandwidth.

[0164] The computer-readable storage medium proposed in this embodiment belongs to the field of artificial intelligence technology and is applied to the network traffic regulation scenario in online services such as medical health or fintech. This application monitors and obtains the real-time traffic characterization data at each network node in the target system within the target time period; performs amplification processing to obtain a batch of sample data; randomly divides it to obtain a training set and a validation set; inputs them into a long short-term memory neural network based on the ant colony algorithm to train an AI intelligent model for real-time network traffic prediction; transmits the real-time traffic at each network node recorded by the real-time traffic monitoring component to the AI intelligent model for real-time network traffic prediction; obtains the network predicted traffic corresponding to each network node at the future target time point; and combines the network traffic bandwidth adjustment strategy to adjust the network traffic bandwidth of each network node at the future target time point. The ant colony algorithm is used for AI intelligent model training, and hyperparameter combination selection is performed to ensure the high availability and prediction accuracy of the trained AI intelligent model.

[0165] Through the description of the above embodiments, those skilled in the art can clearly understand that the above-described embodiment methods can be implemented by means of software plus a necessary general hardware platform. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on such an understanding, the technical solution of the present application, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) and includes several instructions for causing a terminal device (which can be a mobile phone, computer, server, air conditioner, or network device, etc.) to execute the methods described in various embodiments of the present application.

[0166] Obviously, the above-described embodiments are only a part of the embodiments of the present application, rather than all of the embodiments. The accompanying drawings show the preferred embodiments of the present application, but do not limit the patent scope of the present application. The present application can be implemented in many different forms. On the contrary, the purpose of providing these embodiments is to make the understanding of the disclosed content of the present application more thorough and comprehensive. Although the present application has been described in detail with reference to the foregoing embodiments, for those skilled in the art, they can still modify the technical solutions described in the foregoing specific embodiments, or perform equivalent replacements for some of the technical features. Any equivalent structure directly or indirectly using the content of the specification and drawings of the present application in other related technical fields is equally within the scope of the patent protection of the present application. The non-company software tools or components appearing in the embodiments of the present application are only introduced by way of example and do not represent actual use.

Claims

1. A method for real-time dynamic adjustment of network traffic bandwidth, characterized in that: The steps include: Monitor and obtain real-time traffic characterization data at each network node in the target system within the target time period; Amplifying the real-time traffic characterization data at each network node using a unified amplification method to obtain batch sample data; Randomly divide the batch sample data according to a preset ratio value to obtain a training set and a validation set; Inputting the training set and the verification set into a long short-term memory neural network based on an ant colony algorithm to train an AI intelligent model for real-time prediction of network traffic; Establishing a data transmission relationship between the trained AI intelligent model for real-time prediction of network traffic and the real-time traffic monitoring components preset for each network node; Based on the data transmission relationship, the real-time traffic at each network node recorded by the real-time traffic monitoring component is transmitted to the AI ​​intelligent model for real-time prediction of network traffic; Obtain the predicted network traffic corresponding to each network node output by the AI ​​intelligent model at a future target time point; According to the predicted network traffic corresponding to each network node at the future target time point, and the preset network traffic bandwidth adjustment strategy, the network traffic bandwidth of each network node is adjusted at the future target time point.

2. The method for real-time dynamic adjustment of network traffic bandwidth according to claim 1, characterized in that: The step of monitoring and acquiring real-time traffic characterization data at each network node in the target system within the target time period specifically includes: Starting a real-time traffic monitoring component preset for each of the network nodes; According to the real-time traffic monitoring component and the node number information of each network node, record the real-time traffic corresponding to each network node at each unit time point, wherein each unit time point includes every second; Based on the recorded results, draw the real-time traffic change curves corresponding to each of the network nodes; The real-time traffic change curves corresponding to the respective network nodes are set as the real-time traffic characterization data at the respective network nodes.

3. The method for real-time dynamic adjustment of network traffic bandwidth according to claim 1, characterized in that: The step of inputting the training set and the verification set into a long short-term memory neural network based on an ant colony algorithm to train an AI intelligent model for real-time prediction of network traffic specifically includes: Identifying model hyperparameters to be optimized in the long short-term memory neural network, wherein the model hyperparameters to be optimized include the number of hidden units, the learning rate, and the number of training cycles; Generate a target number of hyperparameter combinations using a random combination method and preset model hyperparameter optional values, wherein the preset model hyperparameter optional values ​​include optional values ​​corresponding to the number of hidden units, the learning rate, and the number of training cycles, respectively; The target number of hyperparameter combinations is set as an ant colony in an ant algorithm; Selecting in sequence the hyperparameter combinations corresponding to different ant monomers in the ant colony as training parameters of the long short-term memory neural network, and performing network training in combination with the training set and the validation set; According to the training results, the hyperparameter combination corresponding to the minimum loss value of the verification result corresponding to the verification set is selected as the target hyperparameter combination, wherein the minimum value is less than a preset loss threshold; The long short-term memory neural network trained based on the target hyperparameter combination is used as the AI ​​intelligent model for real-time prediction of network traffic.

4. The method for real-time dynamic adjustment of network traffic bandwidth according to claim 3, characterized in that: The step of sequentially selecting the hyperparameter combinations corresponding to different ant monomers in the ant colony as the training parameters of the long short-term memory neural network, and performing network training in combination with the training set and the verification set specifically includes: Step 701, setting the hyperparameter combination corresponding to the current ant monomer as the training parameters of the long short-term memory neural network to complete the initial deployment of the long short-term memory neural network; Step 702, inputting the training set into the long short-term memory neural network that has been initially deployed, and performing learning training on the long short-term memory neural network; Step 703, until the learning rate corresponding to the long short-term memory neural network reaches the learning rate in the hyperparameter combination and the number of training cycles also reaches the number of training cycles in the hyperparameter combination, stop the learning and training of the long short-term memory neural network to obtain the long short-term memory neural network to be verified; Step 704: input the verification set into the long short-term memory neural network to be verified, and obtain the real-time traffic prediction change curve at each network node in the target system output by the long short-term memory neural network to be verified; Step 705, obtaining the loss value corresponding to the hyperparameter combination according to the real-time traffic prediction change curve at each network node in the target system and the real-time traffic actual change curve at each network node in the validation set; Step 706, taking different ant monomers as current ant monomers in turn, and repeatedly executing steps 701 to 705, until the loss values ​​corresponding to all hyperparameter combinations are obtained, and the network training is completed.

5. The method for real-time dynamic adjustment of network traffic bandwidth according to claim 4, characterized in that: The step of obtaining the loss value corresponding to the hyperparameter combination according to the real-time traffic prediction change curve at each network node in the target system and the real-time traffic actual change curve at each network node in the validation set specifically includes: Identify the time period of the actual change curve of the real-time traffic at each network node in the verification set to obtain the time period corresponding to each real-time traffic change curve; According to the time period corresponding to each real-time traffic change curve and the network node corresponding to each real-time traffic change curve, the real-time traffic prediction change curve corresponding to the corresponding network node in the same time period is obtained from the corresponding real-time traffic prediction change curve; The real-time traffic prediction change curve corresponding to the same network node in the same time period is superimposed and compared with the real-time traffic actual change curve; The unsuperimposed curve parts are marked with distinctive colors, and the proportion of the total superimposed time occupied by the curve parts marked with distinctive colors is finally counted; The total overlap time ratio value is used as the loss value corresponding to the hyperparameter combination.

6. The method for real-time dynamic adjustment of network traffic bandwidth according to claim 1, characterized in that: Before executing the step of transmitting the real-time traffic at each network node recorded by the real-time traffic monitoring component to the AI ​​intelligent model for real-time prediction of network traffic, the method further includes: According to a preset abnormal value detection method, abnormal value detection is performed on the real-time traffic at each network node recorded by the real-time traffic monitoring component to obtain abnormal value detection results, wherein the abnormal values ​​include real-time traffic values ​​with obvious recording errors and obvious omissions; Combining the real-time traffic corresponding to the previous unit time point and the next unit time point of the target abnormal value, each abnormal value in the abnormal value detection result is corrected; The correction processing results are added to the real-time traffic at each network node recorded by the real-time traffic monitoring component to update the recorded results.

7. The method for real-time dynamic adjustment of network traffic bandwidth according to claim 1, characterized in that: The step of adjusting the network traffic bandwidth of each network node at the future target time point according to the network predicted traffic corresponding to each network node at the future target time point and the preset network traffic bandwidth adjustment strategy specifically includes: Identifying the network traffic bandwidth allocated in advance to each of the network nodes as the allocated bandwidth; If the predicted network traffic corresponding to the current network node at the future target time point exceeds the actual traffic allowed to pass by the allocated bandwidth corresponding to the current network node, dynamically expand the allocated bandwidth at the future target time point; If the network predicted traffic corresponding to the current network node at the future target time point does not exceed the actual traffic allowed by the allocated bandwidth corresponding to the current network node, and the network predicted traffic exceeds half of the actual traffic allowed to pass, the allocated bandwidth is still maintained at the future target time point; If the network predicted traffic corresponding to the current network node at the future target time point does not exceed the actual traffic allowed by the allocated bandwidth corresponding to the current network node, and the network predicted traffic does not exceed half of the actual traffic allowed to pass, the allocated bandwidth is dynamically reduced at the future target time point.

8. A real-time dynamic adjustment device for network traffic bandwidth, characterized in that: include: A real-time traffic characterization data acquisition module is used to monitor and acquire real-time traffic characterization data at each network node in the target system within a target time period; A real-time traffic characterization data amplification module, used to amplify the real-time traffic characterization data at each network node in a unified amplification manner to obtain batch sample data; A training set and a validation set partitioning module is used to randomly partition the batch sample data according to a preset ratio value to obtain a training set and a validation set; An AI intelligent model training module, used for inputting the training set and the verification set into a long short-term memory neural network based on an ant colony algorithm to train an AI intelligent model for real-time prediction of network traffic; A data transmission relationship establishment module, used to establish a data transmission relationship between the trained AI intelligent model for real-time prediction of network traffic and the real-time traffic monitoring components preset for each network node; A real-time traffic record result transmission module, used to transmit the real-time traffic at each network node recorded by the real-time traffic monitoring component to the AI ​​intelligent model for real-time prediction of network traffic based on the data transmission relationship; A network prediction traffic output module is used to obtain the network prediction traffic corresponding to each network node output by the AI ​​intelligent model at a future target time point; The network traffic bandwidth adjustment module is used to adjust the network traffic bandwidth of each network node at the future target time point according to the network predicted traffic corresponding to each network node at the future target time point and the preset network traffic bandwidth adjustment strategy.

9. A computer device, characterized in that: The method comprises a memory and a processor, wherein the memory stores computer-readable instructions, and when the processor executes the computer-readable instructions, the steps of the method for real-time dynamic adjustment of network traffic bandwidth as described in any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer-readable instructions, and when the computer-readable instructions are executed by the processor, the steps of the method for real-time dynamic adjustment of network traffic bandwidth according to any one of claims 1 to 7 are implemented.

Images