Network resource scheduling and optimizing method, system and device based on artificial intelligence and medium

Abstract

The invention provides a network resource scheduling and optimizing method, system and device based on artificial intelligence and a medium, and belongs to the technical field of network optimizing.The method comprises the steps that network communication state data are collected and preprocessed; various types of artificial intelligence models are constructed, historical network communication state data are used for training, and the models are embedded into an intelligent scheduling server; the artificial intelligence model comprises a traffic prediction model, a network anomaly detection model and a routing optimization model; the intelligent scheduling server predicts a traffic mode by using a traffic prediction model and detects network anomaly information by using a network anomaly detection model; and the intelligent scheduling server performs routing path optimization by using a routing optimization model, and dynamically allocates network resources in combination with network abnormal information, a routing path optimization result and a predicted flow mode. According to the invention, by dynamically allocating the network resources, intelligent scheduling and optimization of the network resources can be realized, and the utilization rate of the network resources is improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of network optimization, and specifically relates to a network resource scheduling and optimization method, system, equipment and medium based on artificial intelligence. Background Art

[0002] With the rapid development of the Internet, the management and optimization of network resources have become increasingly complex. Especially driven by big data, cloud computing and the Internet of Things, network communication volume has increased dramatically and network topology has become more complex. Traditional network resource scheduling methods have the following problems when dealing with dynamically changing network communication needs: First, traditional network resource scheduling methods often rely on manual configuration and static rules, and cannot respond to changes in network status in real time. For example, when the network is congested, the update speed of manual configuration is far behind the changes in network status, resulting in the inability to adjust resource allocation in time, and thus unable to handle emergencies, causing network performance to decline. Secondly, it is difficult to accurately predict traffic patterns and network demand. For example, in a certain network, due to inaccurate predictions, resources may be insufficient during peak traffic, and resources may be wasted during low traffic, resulting in low utilization of network resources. Thirdly, the network anomaly detection method is single, relying on simple rule matching or threshold judgment, and it is difficult to identify complex network attacks or faults, especially for new network attacks. Traditional detection methods cannot detect them in time, resulting in network security risks. Finally, traditional routing optimization methods are based on fixed routing tables or simple path selection methods, and cannot be intelligently adjusted according to network status and traffic demand. Especially when the network topology changes or traffic is unevenly distributed, the routing path is often not optimal, affecting network performance and resource utilization. Summary of the invention

[0003] In a first aspect, an embodiment of the present application provides a network resource scheduling and optimization method based on artificial intelligence, comprising the following steps: S1. Collect and preprocess network communication status data, the network communication status data includes user communication needs, network topology, channel status, traffic data and traffic pattern; S2. Construct various types of artificial intelligence models, and use historical network communication status data for training, and embed the trained artificial intelligence models of various types into the intelligent scheduling server; the artificial intelligence models include traffic prediction models, network anomaly detection models and routing optimization models; S3. The intelligent scheduling server uses the traffic prediction model to predict the traffic pattern based on the real-time traffic data, and uses the network anomaly detection model to detect network anomaly information based on the real-time traffic data; S4. The intelligent scheduling server uses the routing optimization model to combine the network topology, predicted network anomaly information and predicted traffic patterns to optimize the routing path, and then dynamically allocates network resources based on the network anomaly information, routing path optimization results and predicted traffic patterns.

[0004] Furthermore, the specific steps of step S1 are as follows: S11. Collecting user communication needs through the user terminal device, the user communication needs include the requested service type, data volume and response time requirements; S12. Use a network management tool to obtain network topology and channel status, wherein the network topology includes nodes, links, and connection relationships between nodes through links, and the channel status includes signal strength, bandwidth, and bit error rate; S13. Real-time collection of traffic data by flow detection tools set on network nodes or intelligent scheduling servers, the flow data includes flow size and flow rate at different times and from different sources; S14. Analyze the traffic data using deep packet inspection to identify traffic patterns; S15. Perform data cleaning, normalization and feature extraction preprocessing on the collected user communication requirements, network topology, channel status and traffic patterns.

[0005] Furthermore, the specific steps of step S2 are as follows: S21. Select a time series prediction model as a traffic prediction model, and use historical traffic data for training to adjust the time series prediction model to minimize the prediction error, and deploy the trained time series prediction model to the intelligent scheduling server to predict traffic data and traffic patterns; S22. Select a convolutional neural network model as a network anomaly detection model, use the labeled normal traffic data and abnormal traffic data as training samples to train the abnormal network detection model, and use the cross entropy loss function to optimize the model during the training process, and deploy the trained network anomaly detection model to the intelligent scheduling server for network anomaly detection; S23. Build a routing optimization model based on reinforcement learning algorithm, perform environmental modeling based on network topology, traffic prediction results, traffic pattern, and channel status, initialize the Q value table, select routing actions, complete agent training, and deploy the routing optimization model to the intelligent scheduling server for routing optimization.

[0006] Furthermore, the specific steps of step S21 are as follows: S211. Determine the following ARIMA model as the traffic prediction model: ; in, is the time series of flow data, m is the autoregressive order, n is the moving average order, is the autoregressive coefficient, is the moving average coefficient, i is the index of the autoregressive, j is the index of the moving average, is the white noise at time t; S212. Determine the values ​​of the autoregressive order m and the moving average order n; S213. Use the time series of traffic data as training data, and divide the training data into a training set and a test set according to a set ratio; S214. Use the test set to fit the ARIMA model and estimate model parameters … and … ; S215. According to the model parameters … and … Determine the autoregressive term and moving average term of the ARIMA model; S216. Use the test set and minimize the mean square error to evaluate the prediction performance of the ARIMA model and obtain the final traffic prediction model; S217. Deploy the trained traffic prediction model to the intelligent scheduling server; The specific steps of step S22 are as follows: S221. Select convolutional neural network as network anomaly detection model; S222. Construct a convolutional layer of a convolutional neural network, and characterize the three-dimensional flow data X input to the convolutional layer with the time series l of the flow data, the flow feature v, and the type k of the flow feature, determine the height h, width w, and number c of convolution output channels of the convolution kernel for the convolution operation for feature extraction, and obtain the output expression of the convolutional layer; S223. Construct a pooling layer of a convolutional neural network, select a maximum pooling window, and use the output expression of the convolutional layer as input to calculate the output expression of the pooling layer; S224. construct a fully connected layer of the convolutional neural network, and use the output expression of the pooling layer as input to calculate the output of the fully connected layer, thereby obtaining a convolutional neural network model; S225. Divide the traffic data samples marked with normal and abnormal states into a training set and a test set and perform normalization processing; S226. Initialize the parameters of the convolutional neural network model, set the cross loss function, select the stochastic gradient descent optimizer and the learning rate for iterative training, and adjust the model parameters during the iterative training process to minimize the loss function of the convolutional neural network model on the training set; S227. Use the test set to verify the performance of the convolutional neural network model and obtain the final network anomaly detection model; S228. Deploy the network anomaly detection model to the intelligent scheduling server.

[0007] Furthermore, the specific steps of step S23 are as follows: S231. Perform environmental modeling on the reinforcement learning algorithm model, construct a network node adjacency matrix according to the network topology, set weights for links between network nodes according to signal states, quantify link traffic of different service types according to traffic prediction results and traffic pattern analysis results, obtain link delay time through network detection tools, and adjust link delay properties in combination with channel states; S232. Construct a two-dimensional table, and use rows to represent node states and lists to represent routing actions, set the initial value of the table to 0, and complete the initialization of the Q value table; S233. The action of the agent is defined as a routing decision for forwarding a data packet from the current node to the next node, and the action space is set as the set of all reachable adjacent nodes of the current node in the network, and each adjacent node corresponds to an optional routing forwarding action; S234. Randomly design the routing path of the agent at each network node, and make instant rewards based on the delay reduction and traffic balance factors caused by the routing path, and then update the Q value table based on the instant rewards; S235. Selecting a routing forwarding action in the action space content to iteratively train the agent until the Q value table converges and the probability of the selected routing path similarity is greater than a threshold, thereby obtaining a final routing optimization model; S236. Deploy the routing optimization model to the intelligent scheduling server, and the intelligent scheduling server sends a routing table to each network node.

[0008] Furthermore, the specific steps of step S3 are as follows: S31. The intelligent dispatch server inputs real-time traffic data into the traffic prediction model to predict traffic data and predict the flow size and flow rate of various types of data; S32. The intelligent scheduling server inputs the real-time traffic data into the network anomaly detection model to identify whether there is a network anomaly, and determines the type of network anomaly and the link location of the network anomaly.

[0009] Further, step S4 is specifically as follows: S41. The intelligent scheduling server obtains the predicted traffic data and network anomaly information, adjusts the Q value table of the routing optimization model in combination with the real-time network topology until the Q value table converges, and uses the path with a high reward score in the optimized Q value table to adjust the routing to obtain the adjusted routing path; S42. The intelligent scheduling server allocates bandwidth resources for each link based on the adjusted routing path, combined with the predicted network anomaly information and predicted traffic patterns, and allocates computing resources and cache space to network nodes based on load and business needs; S43. The intelligent scheduling server sends the adjusted routing path and resource allocation plan to the network nodes, while monitoring the network communication status.

[0010] In a second aspect, the embodiment of the present application further provides a network resource scheduling and optimization system based on artificial intelligence, including: A data acquisition module is used to collect and pre-process network communication status data, wherein the network communication status data includes user communication requirements, network topology, channel status, flow data and flow pattern; Artificial intelligence model construction and training module, used to build various types of artificial intelligence models, and use historical network communication status data for training, and embed the trained artificial intelligence models of various types into the intelligent scheduling server; the artificial intelligence models include traffic prediction models, network anomaly detection models and routing optimization models; The traffic prediction and network anomaly detection module is used to use the traffic prediction model in the intelligent scheduling server to predict the traffic pattern according to the real-time traffic data, and use the network anomaly detection model to detect network anomaly information according to the real-time traffic data; The routing optimization and network resource allocation module is used to use the routing optimization model in the intelligent scheduling server to combine the network topology, network anomaly information and predicted traffic patterns to optimize the routing path, and then dynamically allocate network resources in combination with the network anomaly information, routing path optimization results and predicted traffic patterns.

[0011] In a third aspect, an embodiment of the present application further provides an electronic device, comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, wherein when the processor executes the program, the steps of the network resource scheduling and optimization method based on artificial intelligence as described in the first aspect are implemented.

[0012] In a fourth aspect, an embodiment of the present application further provides a storage medium on which a computer program is stored, and when the computer program is executed by a processor, the steps of the network resource scheduling and optimization method based on artificial intelligence as described in the first aspect are implemented.

[0013] It can be seen from the above technical solutions that the present invention has the following advantages: The artificial intelligence-based network resource scheduling and optimization method, system, device and medium provided in the present application realizes the process of network resource scheduling and optimization based on artificial intelligence technology. By collecting network communication status data and performing preprocessing, various types of artificial intelligence models are constructed and trained, so that the intelligent scheduling server can predict traffic patterns in real time, detect network anomalies, and optimize routing paths according to network topology and traffic patterns to realize dynamic allocation of resources, thereby improving the overall utilization of network resources, ensuring network safe operation, improving network communication efficiency, meeting the diverse needs of users, and being able to cope with various complex network environments and business needs. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] In order to more clearly illustrate the technical solution of the present invention, the accompanying drawings required for use in the description will be briefly introduced below. Obviously, the accompanying drawings in the following description are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.

[0015] Figure 1 It is a flow chart of the network resource scheduling and optimization method based on artificial intelligence of the present invention.

[0016] Figure 2 Schematic diagram of the network resource scheduling and optimization system based on artificial intelligence of the present invention. DETAILED DESCRIPTION

[0017] In the specific steps of the network resource scheduling and optimization method based on artificial intelligence, various embodiments of the present disclosure will be described in detail below. The present disclosure may have various embodiments, and adjustments and changes may be made therein. However, it should be understood that there is no intention to limit the various embodiments of the present disclosure to the specific embodiments disclosed herein, but the present disclosure should be understood to cover all adjustments, equivalents and / or alternatives that fall within the spirit and scope of the various embodiments of the present disclosure.

[0018] For example, with the rapid development of the Internet, the management and optimization of network resources are facing unprecedented complexity. Especially under the strong promotion of big data, cloud computing and Internet of Things technologies, network communication volume has risen sharply and network topology has become increasingly complex. When dealing with the ever-changing network communication needs, traditional network resource scheduling strategies have the following drawbacks: First, most of these traditional methods rely on manual configuration and static rules, and it is difficult to respond quickly to real-time changes in network status. Taking network congestion as an example, the update speed of manual configuration lags far behind the changes in network status, resulting in the inability to adjust resource allocation in time and effectively respond to emergencies, which in turn causes a decline in network performance. Secondly, accurate prediction of traffic patterns and network demand has become a major problem. Taking a certain network as an example, due to inaccurate predictions, resource shortages may occur during peak traffic periods, while resources may be idle during low periods, greatly reducing the utilization efficiency of network resources. Furthermore, the means of network anomaly detection are relatively simple, mainly relying on simple rule matching or threshold judgment, and it is difficult to effectively identify complex network attacks or faults. Especially for new network attacks, traditional detection methods are often difficult to detect in time, which brings huge hidden dangers to network security. Finally, traditional routing optimization strategies are based on fixed routing tables or simple path selection methods, and cannot be intelligently adjusted according to network status and traffic demand. When the network topology changes or traffic is unevenly distributed, the routing path is often not optimal, which affects network performance and resource utilization.

[0019] In response to the above problems, this embodiment provides a network resource scheduling and optimization method based on artificial intelligence. By collecting network communication status data, building and training multiple artificial intelligence models, intelligent scheduling and optimization of network resources are achieved.

[0020] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0021] See also Figure 1 The figure is a flowchart of a network resource scheduling and optimization method based on artificial intelligence in a specific embodiment, and the method includes the following steps: S1. Collect and preprocess network communication status data, the network communication status data includes user communication needs, network topology, channel status, traffic data and traffic pattern; It should be noted that by collecting network communication status data and preprocessing it, comprehensive network communication status information can be obtained, providing a basis for subsequent artificial intelligence model training and optimization. Preprocessing can also reduce the impact of noise and redundant information on model performance. S2. Construct various types of artificial intelligence models, and use historical network communication status data for training, and embed the trained artificial intelligence models of various types into the intelligent scheduling server; the artificial intelligence models include traffic prediction models, network anomaly detection models and routing optimization models; It should be noted that by building various types of artificial intelligence models and using historical network communication status data for training, it is possible to achieve intelligent analysis and prediction of network communication status, and the trained models can be embedded in the intelligent scheduling server to provide a basis for subsequent traffic prediction, network anomaly detection and routing optimization; S3. The intelligent scheduling server uses the traffic prediction model to predict the traffic pattern based on the real-time traffic data, and uses the network anomaly detection model to detect network anomaly information based on the real-time traffic data; It should be noted that by using traffic prediction models and network anomaly detection models for real-time analysis and detection, it is possible to accurately predict network traffic and promptly discover network anomalies, providing a decision-making basis for subsequent routing optimization and resource allocation; S4. The intelligent scheduling server uses the routing optimization model to combine the network topology, predicted network anomaly information and predicted traffic patterns to optimize the routing path, and then dynamically allocates network resources based on the network anomaly information, routing path optimization results and predicted traffic patterns; It should be noted that by dynamically allocating network resources, intelligent scheduling and optimization of network resources can be achieved, thereby improving the utilization rate of network resources and the stability of the network.

[0022] This embodiment collects network communication status data, builds and trains multiple artificial intelligence models, and realizes intelligent scheduling and optimization of network resources. It can dynamically adapt to network changes, improve network resource utilization, reduce the impact of network anomalies on business, and thus improve user experience.

[0023] Further, as a refinement and extension of the specific implementation of the above embodiment, in order to fully illustrate the specific implementation process in this embodiment, another network resource scheduling and optimization method based on artificial intelligence is provided, and the method includes the following steps: S1. Collect and pre-process network communication status data, the network communication status data includes user communication needs, network topology, channel status, traffic data and traffic mode; the specific steps of step S1 are as follows: S11. Collecting user communication needs through the user terminal device, the user communication needs include the requested service type, data volume and response time requirements; Exemplarily, user communication requirements may include request type and response time, network topology is the connection relationship of network devices, and channel status may include packet loss rate and data delay; S12. Use a network management tool to obtain network topology and channel status, wherein the network topology includes nodes, links, and connection relationships between nodes through links, and the channel status includes signal strength, bandwidth, and bit error rate; S13. Real-time collection of traffic data by flow detection tools set on network nodes or intelligent scheduling servers, the flow data includes flow size and flow rate at different times and from different sources; S14. Analyze network traffic data using deep packet inspection to identify traffic patterns; Exemplarily, the traffic pattern may include peak periods and trough periods; S15. Perform data cleaning, normalization and feature extraction preprocessing on the collected user communication requirements, network topology, channel status and traffic patterns; Exemplarily, data cleaning includes removing duplicate data, filling missing data, and filtering invalid data; S2. Construct various types of artificial intelligence models, and use historical network communication status data for training, and embed the trained artificial intelligence models of various types into the intelligent scheduling server; the artificial intelligence models include traffic prediction models, network anomaly detection models and routing optimization models; the specific steps of step S2 are as follows: S21. Select a time series prediction model as the traffic prediction model, and use historical traffic data for training to adjust the time series prediction model to minimize the prediction error, and deploy the trained time series prediction model to the intelligent scheduling server to predict traffic data and traffic patterns; the specific steps of step S21 are as follows: S211. Determine the following ARIMA model as the traffic prediction model: ; in, is the time series of flow data, m is the autoregressive order, n is the moving average order, is the autoregressive coefficient, is the moving average coefficient, i is the index of the autoregressive, j is the index of the moving average, is the white noise at time t; S212. Determine the values ​​of the autoregressive order m and the moving average order n; S213. Use the time series of traffic data as training data, and divide the training data into a training set and a test set according to a set ratio; S214. Use the test set to fit the ARIMA model and estimate model parameters … and … ; S215. According to the model parameters … and … Determine the autoregressive term and moving average term of the ARIMA model; S216. Use the test set and minimize the mean square error to evaluate the prediction performance of the ARIMA model and obtain the final traffic prediction model; S217. Deploy the trained traffic prediction model to the intelligent scheduling server; S22. Select a convolutional neural network model as a network anomaly detection model, use the marked normal traffic data and abnormal traffic data as training samples to train the abnormal network detection model, and use the cross entropy loss function to optimize the model during the training process, and deploy the trained network anomaly detection model to the intelligent scheduling server for network anomaly detection; the specific steps of step S22 are as follows: S221. Select convolutional neural network as network anomaly detection model; S222. Construct a convolutional layer of a convolutional neural network, and characterize the three-dimensional flow data X input to the convolutional layer with the time series l of the flow data, the flow feature v, and the type k of the flow feature, determine the height h, width w, and number c of convolution output channels of the convolution kernel for the convolution operation for feature extraction, and obtain the output expression of the convolutional layer; S223. Construct a pooling layer of a convolutional neural network, select a maximum pooling window, and use the output expression of the convolutional layer as input to calculate the output expression of the pooling layer; S224. construct a fully connected layer of the convolutional neural network, and use the output expression of the pooling layer as input to calculate the output of the fully connected layer, thereby obtaining a convolutional neural network model; S225. Divide the traffic data samples marked with normal and abnormal states into a training set and a test set and perform normalization processing; S226. Initialize the parameters of the convolutional neural network model, set the cross loss function, select the stochastic gradient descent optimizer and the learning rate for iterative training, and adjust the model parameters during the iterative training process to minimize the loss function of the convolutional neural network model on the training set; S227. Use the test set to verify the performance of the convolutional neural network model and obtain the final network anomaly detection model; S228. Deploy the network anomaly detection model to the intelligent scheduling server; S23. Construct a routing optimization model based on a reinforcement learning algorithm, perform environmental modeling based on network topology, traffic prediction results, traffic patterns, and channel states, initialize the Q value table, select routing actions, complete agent training, and deploy the routing optimization model to the intelligent scheduling server for routing optimization; the specific steps of step S23 are as follows: S231. Perform environmental modeling on the reinforcement learning algorithm model, construct a network node adjacency matrix according to the network topology, set weights for links between network nodes according to signal states, quantify link traffic of different service types according to traffic prediction results and traffic pattern analysis results, obtain link delay time through network detection tools, and adjust link delay properties in combination with channel states; Exemplarily, an adjacency matrix is ​​constructed to represent the connection relationship between nodes. For example, network node A is connected to network node B and network node C, and network node B is connected to network node A and network node D. The weight is set according to the link bandwidth. If the link bandwidth of node AB is 10Mbps, the link bandwidth of network node AC is 5Mbps, and the link bandwidth of network node BD is 8Mbps, the link bandwidth with a higher value can be set with a higher weight, and the weight can be set to the link bandwidth value; if the video stream type traffic accounts for 60% of the link and the file transfer type traffic accounts for 40%, and according to the traffic prediction model, at a certain moment, the traffic of network node A is 2Mbps, of which the video stream type traffic is 1.2Mbps and the file transfer type traffic is 0.8Mbps; the link round-trip delay time is obtained through the network detection tool. If the link round-trip delay time of the network node AB link is 20ms, the delay attribute is adjusted in combination with factors such as signal strength and bit error rate in the channel state. For example, if the signal strength is -60dBm and the bit error rate is 0.1%, the delay attribute can be adjusted to a higher value, such as 0.8; S232. Construct a two-dimensional table, and use rows to represent node states and lists to represent routing actions, set the initial value of the table to 0, and complete the initialization of the Q value table; S233. The action of the agent is defined as a routing decision for forwarding a data packet from the current node to the next node, and the action space is set as the set of all reachable adjacent nodes of the current node in the network, and each adjacent node corresponds to an optional routing forwarding action; S234. Randomly design the routing path of the agent at each network node, and give instant rewards based on the delay reduction and traffic balance factors caused by the routing path, and then update the Q value table using the following formula based on the instant rewards;

[0024] in, is the current node status, is the current routing action taken, Is to perform routing actions After receiving the instant reward, is a discount factor used to measure the importance of future rewards, and its value range is usually between 0 and 1. The learning rate is used to control the degree of mixing between the previous value and the current iteration value in each iteration. The value range is usually between 0 and 1. Is to perform routing actions The next node state entered after, maxQ(s',a') represents the next node state All possible routing actions The largest Q value in S235. Selecting a routing forwarding action in the action space content to iteratively train the agent until the Q value table converges and the probability of the selected routing path similarity is greater than a threshold, thereby obtaining a final routing optimization model; For example, in a network routing scenario, the agent selects a routing path for a data packet from a source node to a destination node according to the Q value table; if after a certain number of iterations, the agent always selects the same or several similar routing paths, for example, the probability of selecting a certain path exceeds 90%, and the routing strategy is no longer frequently changed, then it indicates that the strategy provided by the Q value table is relatively stable, and the agent has found a better routing strategy in the current environment; S236. Deploy the routing optimization model to the intelligent scheduling server, and the intelligent scheduling server sends a routing table to each network node; S3. The intelligent scheduling server uses a traffic prediction model to predict traffic patterns based on real-time traffic data, and uses a network anomaly detection model to detect network anomaly information based on real-time traffic data; the specific steps of step S3 are as follows: S31. The intelligent dispatch server inputs real-time traffic data into the traffic prediction model to predict traffic data and predict the flow size and flow rate of various types of data; S32. The intelligent scheduling server inputs the real-time traffic data into the network anomaly detection model to identify whether there is a network anomaly, and determines the type of network anomaly and the link location of the network anomaly; S4. The intelligent scheduling server uses the routing optimization model to combine the network topology, predicted network anomaly information and predicted traffic patterns to optimize the routing path, and then dynamically allocates network resources in combination with the network anomaly information, routing path optimization results and predicted traffic patterns; Step S4 is as follows: S41. The intelligent scheduling server obtains the predicted traffic data and network anomaly information, adjusts the Q value table of the routing optimization model in combination with the real-time network topology until the Q value table converges, and uses the path with a high reward score in the optimized Q value table to adjust the routing to obtain the adjusted routing path; S42. The intelligent scheduling server allocates bandwidth resources for each link based on the adjusted routing path, combined with the predicted network anomaly information and predicted traffic patterns, and allocates computing resources and cache space to network nodes based on load and business needs; S43. The intelligent scheduling server sends the adjusted routing path and resource allocation plan to the network nodes, while monitoring the network communication status.

[0025] It should be understood that the order of execution of the steps in the above embodiment does not necessarily mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiment of the present invention.

[0026] like Figure 2 As shown, the following is an embodiment of the network resource scheduling and optimization system based on artificial intelligence provided by the embodiment of the present disclosure. The system and the network resource scheduling and optimization methods based on artificial intelligence in the above-mentioned embodiments belong to the same inventive concept. For details not described in detail in the embodiment of the network resource scheduling and optimization system based on artificial intelligence, please refer to the embodiment of the network resource scheduling and optimization method based on artificial intelligence.

[0027] The system includes: A data acquisition module is used to collect and pre-process network communication status data, wherein the network communication status data includes user communication requirements, network topology, channel status, flow data and flow pattern; Artificial intelligence model construction and training module, used to build various types of artificial intelligence models, and use historical network communication status data for training, and embed the trained artificial intelligence models of various types into the intelligent scheduling server; the artificial intelligence models include traffic prediction models, network anomaly detection models and routing optimization models; The traffic prediction and network anomaly detection module is used to use the traffic prediction model in the intelligent scheduling server to predict the traffic pattern according to the real-time traffic data, and use the network anomaly detection model to detect network anomaly information according to the real-time traffic data; The routing optimization and network resource allocation module is used to use the routing optimization model in the intelligent scheduling server to combine the network topology, network anomaly information and predicted traffic patterns to optimize the routing path, and then dynamically allocate network resources in combination with the network anomaly information, routing path optimization results and predicted traffic patterns.

[0028] This embodiment uses a data acquisition module, an artificial intelligence model construction and training module, a traffic prediction and network anomaly detection module, and a routing optimization and network resource allocation module to achieve comprehensive collection and analysis of network communication status, as well as intelligent scheduling and optimization of network resources, thereby improving the overall performance of network communication and user experience.

[0029] The network resource scheduling and optimization method based on artificial intelligence provided in the embodiment of the present application can be applied to electronic devices. It can be understood by those skilled in the art that the electronic device structure involved in the embodiment of the present invention does not constitute a limitation on the electronic device, and the electronic device may include more or less components than shown, or combine certain components, or arrange different components. In an embodiment of the present invention, the electronic device includes but is not limited to a laptop computer, a desktop computer, a workbench, a personal digital assistant, a server, a blade server, a mainframe computer, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples, and are not intended to limit the implementation of the embodiments of the present application described and / or required herein.

[0030] The electronic device may include a processor, an external memory interface, an internal memory, a universal serial bus (USB) interface, a charging management module, a power management module, a battery, a wireless communication module, an audio module, a speaker, a microphone, a sensor module, buttons, a camera, a display, and a SIM card interface, etc.

[0031] It is to be understood that the structure illustrated in the embodiments of the present application does not constitute a specific limitation on the electronic device. In other embodiments of the present application, the electronic device may include more or fewer components than shown in the figure, or combine certain components, or split certain components, or arrange the components differently. The components shown in the figure may be implemented in hardware, software, or a combination of software and hardware.

[0032] The processor may include one or more processing units, for example, the processor may include a central processing unit (CPU), an application processor (AP), a modem processor, a graphics processing unit (GPU), an image signal processor (ISP), a controller, a memory, a video codec, a digital signal processor (DSP), a baseband processor, and / or a neural-network processing unit (NPU), etc. Among them, different processing units may be independent devices or integrated into one or more processors.

[0033] The processor can be the nerve center and command center of the electronic device. The controller can generate an operation control signal according to the instruction operation code and timing signal to complete the control of fetching and executing instructions.

[0034] A memory may also be provided in the processor for storing instructions and data. In some embodiments, the memory in the processor is a cache memory. The memory may store instructions or data that the processor has just used or is cyclically used. If the processor needs to use the instruction or data again, it may be directly called from the memory. This avoids repeated access, reduces the waiting time of the processor, and thus improves system efficiency.

[0035] The above-mentioned electronic device implements the network resource scheduling and optimization method based on artificial intelligence of the present application by collecting network communication status data, building and training multiple artificial intelligence models, and realizing the technical solution of intelligent scheduling and optimization of network resources, so as to achieve the ability to dynamically adapt to network changes, improve network resource utilization, and reduce the impact of network anomalies on business.

[0036] The storage medium provided in the present application stores a program product that can implement an artificial intelligence-based network resource scheduling and optimization method.

[0037] The network resource scheduling and optimization method based on artificial intelligence includes: collecting and preprocessing network communication status data, wherein the network communication status data includes user communication needs, network topology, channel status, traffic data and traffic patterns; constructing various types of artificial intelligence models, and using historical network communication status data for training, and embedding the trained artificial intelligence models of various types into an intelligent scheduling server; the artificial intelligence models include traffic prediction models, network anomaly detection models and routing optimization models; the intelligent scheduling server uses the traffic prediction model to predict traffic patterns based on real-time traffic data, and uses the network anomaly detection model to detect network anomaly information based on real-time traffic data; the intelligent scheduling server uses the routing optimization model to combine the network topology, predicted network anomaly information and predicted traffic patterns to optimize the routing path, and then combines the network anomaly information, routing path optimization results and predicted traffic patterns to dynamically allocate network resources.

[0038] In some possible implementations, the artificial intelligence-based network resource scheduling and optimization method disclosed herein can be implemented in the form of a program product, which includes a program code. When the program product is run on a terminal device, the program code is used to enable the terminal device to execute the steps described in the above "Exemplary Method" section of this specification according to various exemplary implementations of the present disclosure.

[0039] The storage medium of the present disclosure can adopt any combination of one or more readable media. The readable medium can be a readable signal medium or a readable storage medium. The readable storage medium can be, for example, but not limited to, a system, device or device of electricity, magnetism, light, electromagnetic, infrared, or semiconductor, or any combination of the above. More specific examples (non-exhaustive list) of readable storage media include: an electrical connection with one or more wires, a portable disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above.

[0040] The above description of the disclosed embodiments enables one skilled in the art to implement or use the present invention. Various modifications to these embodiments will be apparent to one skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but rather to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A network resource scheduling and optimization method based on artificial intelligence, characterized in that: The steps include: S1. Collect and preprocess network communication status data, the network communication status data includes user communication needs, network topology, channel status, traffic data and traffic pattern; S2. Construct various types of artificial intelligence models, and use historical network communication status data for training, and embed the trained artificial intelligence models of various types into the intelligent scheduling server; the artificial intelligence models include traffic prediction models, network anomaly detection models and routing optimization models; S3. The intelligent scheduling server uses the traffic prediction model to predict the traffic pattern based on the real-time traffic data, and uses the network anomaly detection model to detect network anomaly information based on the real-time traffic data; S4. The intelligent scheduling server uses the routing optimization model to combine the network topology, predicted network anomaly information and predicted traffic patterns to optimize the routing path, and then dynamically allocates network resources based on the network anomaly information, routing path optimization results and predicted traffic patterns.

2. The network resource scheduling and optimization method based on artificial intelligence according to claim 1 is characterized in that: The specific steps of step S1 are as follows: S11. Collecting user communication needs through the user terminal device, the user communication needs include the requested service type, data volume and response time requirements; S12. Use a network management tool to obtain network topology and channel status, wherein the network topology includes nodes, links, and connection relationships between nodes through links, and the channel status includes signal strength, bandwidth, and bit error rate; S13. Real-time collection of traffic data by flow detection tools set on network nodes or intelligent scheduling servers, the flow data includes flow size and flow rate at different times and from different sources; S14. Analyze the traffic data using deep packet inspection to identify traffic patterns; S15. Perform data cleaning, normalization and feature extraction preprocessing on the collected user communication requirements, network topology, channel status and traffic patterns.

3. The network resource scheduling and optimization method based on artificial intelligence according to claim 2 is characterized in that: The specific steps of step S2 are as follows: S21. Select a time series prediction model as a traffic prediction model, and use historical traffic data for training to adjust the time series prediction model to minimize the prediction error, and deploy the trained time series prediction model to the intelligent scheduling server to predict traffic data and traffic patterns; S22. Select a convolutional neural network model as a network anomaly detection model, use the labeled normal traffic data and abnormal traffic data as training samples to train the abnormal network detection model, and use the cross entropy loss function to optimize the model during the training process, and deploy the trained network anomaly detection model to the intelligent scheduling server for network anomaly detection; S23. Build a routing optimization model based on reinforcement learning algorithm, perform environmental modeling based on network topology, traffic prediction results, traffic pattern, and channel status, initialize the Q value table, select routing actions, complete agent training, and deploy the routing optimization model to the intelligent scheduling server for routing optimization.

4. The network resource scheduling and optimization method based on artificial intelligence according to claim 3 is characterized in that: The specific steps of step S21 are as follows: S211. Determine the following ARIMA model as the traffic prediction model: ； in, is the time series of flow data, m is the autoregressive order, n is the moving average order, is the autoregressive coefficient, is the moving average coefficient, i is the index of the autoregressive, j is the index of the moving average, is the white noise at time t; S212. Determine the values ​​of the autoregressive order m and the moving average order n; S213. Use the time series of traffic data as training data, and divide the training data into a training set and a test set according to a set ratio; S214. Use the test set to fit the ARIMA model and estimate model parameters … and … ; S215. According to the model parameters … and … Determine the autoregressive term and moving average term of the ARIMA model; S216. Use the test set and minimize the mean square error to evaluate the prediction performance of the ARIMA model and obtain the final traffic prediction model; S217. Deploy the trained traffic prediction model to the intelligent scheduling server; The specific steps of step S22 are as follows: S221. Select convolutional neural network as network anomaly detection model; S222. Construct a convolutional layer of a convolutional neural network, and characterize the three-dimensional flow data X input to the convolutional layer with the time series l of the flow data, the flow feature v, and the type k of the flow feature, determine the height h, width w, and number c of convolution output channels of the convolution kernel for the convolution operation for feature extraction, and obtain the output expression of the convolutional layer; S223. Construct a pooling layer of a convolutional neural network, select a maximum pooling window, and use the output expression of the convolutional layer as input to calculate the output expression of the pooling layer; S224. construct a fully connected layer of the convolutional neural network, and use the output expression of the pooling layer as input to calculate the output of the fully connected layer, thereby obtaining a convolutional neural network model; S225. Divide the traffic data samples marked with normal and abnormal states into a training set and a test set and perform normalization processing; S226. Initialize the parameters of the convolutional neural network model, set the cross loss function, select the stochastic gradient descent optimizer and the learning rate for iterative training, and adjust the model parameters during the iterative training process to minimize the loss function of the convolutional neural network model on the training set; S227. Use the test set to verify the performance of the convolutional neural network model and obtain the final network anomaly detection model; S228. Deploy the network anomaly detection model to the intelligent scheduling server.

5. The network resource scheduling and optimization method based on artificial intelligence according to claim 4 is characterized in that: The specific steps of step S23 are as follows: S231. Perform environmental modeling on the reinforcement learning algorithm model, construct a network node adjacency matrix according to the network topology, set weights for links between network nodes according to signal states, quantify link traffic of different service types according to traffic prediction results and traffic pattern analysis results, obtain link delay time through network detection tools, and adjust link delay properties in combination with channel states; S232. Construct a two-dimensional table, and use rows to represent node states and lists to represent routing actions, set the initial value of the table to 0, and complete the initialization of the Q value table; S233. The action of the agent is defined as a routing decision for forwarding a data packet from the current node to the next node, and the action space is set as the set of all reachable adjacent nodes of the current node in the network, and each adjacent node corresponds to an optional routing forwarding action; S234. Randomly design the routing path of the agent at each network node, and make instant rewards based on the delay reduction and traffic balance factors caused by the routing path, and then update the Q value table based on the instant rewards; S235. Selecting a routing forwarding action in the action space content to iteratively train the agent until the Q value table converges and the probability of the selected routing path similarity is greater than a threshold, thereby obtaining a final routing optimization model; S236. Deploy the routing optimization model to the intelligent scheduling server, and the intelligent scheduling server sends a routing table to each network node.

6. The network resource scheduling and optimization method based on artificial intelligence according to claim 3 is characterized in that: The specific steps of step S3 are as follows: S31. The intelligent dispatch server inputs real-time traffic data into the traffic prediction model to predict traffic data and predict the flow size and flow rate of various types of data; S32. The intelligent scheduling server inputs the real-time traffic data into the network anomaly detection model to identify whether there is a network anomaly, and determines the type of network anomaly and the link location of the network anomaly.

7. The network resource scheduling and optimization method based on artificial intelligence according to claim 6 is characterized in that: Step S4 is specifically as follows: S41. The intelligent scheduling server obtains the predicted traffic data and network anomaly information, adjusts the Q value table of the routing optimization model in combination with the real-time network topology until the Q value table converges, and uses the path with a high reward score in the optimized Q value table to adjust the routing to obtain the adjusted routing path; S42. The intelligent scheduling server allocates bandwidth resources for each link based on the adjusted routing path, combined with the predicted network anomaly information and predicted traffic patterns, and allocates computing resources and cache space to network nodes based on load and business needs; S43. The intelligent scheduling server sends the adjusted routing path and resource allocation plan to the network nodes, while monitoring the network communication status.

8. A network resource scheduling and optimization system based on artificial intelligence, characterized in that: include: A data acquisition module is used to collect and pre-process network communication status data, wherein the network communication status data includes user communication requirements, network topology, channel status, flow data and flow pattern; Artificial intelligence model construction and training module, used to build various types of artificial intelligence models, and use historical network communication status data for training, and embed the trained artificial intelligence models of various types into the intelligent scheduling server; the artificial intelligence models include traffic prediction models, network anomaly detection models and routing optimization models; The traffic prediction and network anomaly detection module is used to use the traffic prediction model in the intelligent scheduling server to predict the traffic pattern according to the real-time traffic data, and use the network anomaly detection model to detect network anomaly information according to the real-time traffic data; The routing optimization and network resource allocation module is used to use the routing optimization model in the intelligent scheduling server to combine the network topology, network anomaly information and predicted traffic patterns to optimize the routing path, and then dynamically allocate network resources in combination with the network anomaly information, routing path optimization results and predicted traffic patterns.

9. An electronic device, characterized in that: It includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the steps of the network resource scheduling and optimization method based on artificial intelligence as described in any one of claims 1 to 7 are implemented.

10. A storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the network resource scheduling and optimization method based on artificial intelligence as described in any one of claims 1 to 7 are implemented.

Images