A 5G intelligent networking method and system based on edge computing

Abstract

The present invention provides a 5G intelligent networking method and system based on edge computing. The method includes the following steps: deploying edge computing nodes, designing a hierarchical network architecture and a dynamic routing mechanism; deploying an edge model on the edge computing nodes for real-time processing and anomaly detection; dynamically allocating network resources and using the edge model for network switching and load balancing; decomposing complex tasks into multiple subtasks and performing the allocation and parallel processing of the subtasks; updating and optimizing the edge model and dynamically adjusting the networking strategy; data security protection, and monitoring network data and user feedback. The system includes an edge deployment module, a data analysis module, a networking strategy module, a collaborative computing module, a collaborative optimization module, and a security protection module. The present invention can efficiently utilize network resources, perform intelligent networking and network switching, achieve load balancing, greatly improve the network response speed, processing capacity, and adaptability, and enhance the user experience.

Description

Technical Field

[0001] The present invention relates to the technical field of network management, and particularly to a 5G intelligent networking method and system based on edge computing. Background Art

[0002] The configuration, optimization, and management of a network refer to the planning, adjustment, and maintenance of network resources and services to ensure the efficient operation of the network and meet user requirements. This includes setting up network devices, allocating network resources, monitoring network performance, adjusting network parameters, etc. With the popularization of 5G communication technology and the rapid development of the Internet of Things (IoT), the network needs to process a huge amount of data and support diverse application scenarios. Intelligent networking based on the 5G network can utilize the high bandwidth, low latency, and large-scale connection capabilities of the 5G network for automated network management. However, existing networking methods require complex configuration and management, increasing the operation and maintenance difficulty, and the resource allocation is not intelligent enough, resulting in resource waste or insufficiency, further reducing the response speed and affecting network usage. Summary of the Invention

[0003] In order to overcome the deficiencies of the prior art, the purpose of the present invention is to provide a 5G intelligent networking method and system based on edge computing, which can efficiently utilize network resources, perform intelligent networking and network switching, achieve load balancing, greatly improve the network's response speed, processing ability, and adaptability, and enhance the user experience.

[0004] To achieve the above purpose, the present invention provides the following solution: A 5G intelligent networking method based on edge computing, comprising the following steps:

[0005] Select a user area and network nodes, deploy multiple edge computing nodes at the selected location, and design a hierarchical network architecture and a dynamic routing mechanism;

[0006] Collect multi-source data, and deploy a lightweight edge model on the edge computing nodes to perform real-time processing and anomaly detection tasks on the multi-source data, obtaining data analysis results;

[0007] Dynamically allocate network resources according to the data analysis results, and use the edge model for network switching and load balancing;

[0008] Identify complex tasks in the edge model, decompose the complex tasks into multiple subtasks, and allocate the multiple subtasks to different edge computing nodes for parallel processing;

[0009] According to the computing process of the edge computing nodes, perform real-time update and optimization of the edge model, and perform dynamic adjustment of the networking strategy;

[0010] Introduce encryption technology and access control policies during data transmission to protect data security, and continuously monitor network data and user feedback.

[0011] Optionally, select user areas and network nodes, and deploy multiple edge computing nodes at the selected locations, including:

[0012] Through user devices, obtain user traffic data and user geographical information, use big data analysis technology to analyze the obtained historical data, identify user-dense areas, and based on the user-dense areas, identify key nodes in the network;

[0013] Deploy multiple edge computing nodes in the user-dense areas and the key nodes;

[0014] Use virtualization technology to dynamically manage the computing resources of the edge computing nodes, and introduce a resource scheduling system to intelligently allocate computing resources according to the real-time load and task requirements of the edge computing nodes;

[0015] Design a hierarchical network architecture, and deploy an AI-driven routing algorithm in the hierarchical network architecture to perform network load analysis, data path management, and edge node adjustment;

[0016] Among them, the hierarchical network architecture includes a core network layer for overall data management and long-distance data transmission, an edge network layer for local data processing and short-distance transmission, and an access network layer for providing user device access services.

[0017] Optionally, collect multi-source data, and deploy a lightweight edge model on the edge computing nodes to perform real-time processing and anomaly detection tasks on the multi-source data, and obtain data analysis results, including:

[0018] Obtain user device data, base station data, and network device data to get multi-source data;

[0019] According to the time dimension and space dimension, use timestamps and geographical location tags to integrate the multi-source data, and then use semantic technology to analyze and fuse the integrated data to obtain a unified data set;

[0020] Select a lightweight deep learning model, use the unified data set to train the deep learning model, and set anomaly detection rules in the deep learning model to obtain an edge model;

[0021] Deploy the edge model to the edge computing nodes to process and detect anomalies in real-time data, and obtain data analysis results.

[0022] Optionally, the user equipment data includes signal strength, location information, connection time, and application usage, the base station data includes base station load, signal coverage, current active connection count, and available bandwidth, and the network equipment data includes router traffic statistics, switch traffic statistics, latency, and packet loss rate.

[0023] Optionally, dynamically allocate network resources according to the data analysis results, and perform network switching and load balancing using the edge model, including:

[0024] According to the data analysis results, monitor network usage and user requirements in real time, identify the usage trends of bandwidth and spectrum, and use the edge model to predict changes in network requirements, and integrate the identification results and prediction results into resource data;

[0025] According to the resource data, perform dynamic adjustment and allocation of resources in units of the edge computing nodes;

[0026] Predict the user's movement path and movement speed using the user's historical movement data to obtain user travel data, and according to the user travel data, predict a new site or network node before the user leaves the current base station range, and perform network switching;

[0027] Deploy a real-time monitoring dashboard for displaying the current load status and historical data analysis on the edge computing nodes, and deploy an alarm system for load alarms;

[0028] Deploy a load balancing algorithm on the edge computing nodes, and according to the monitoring and alarm results of the monitoring dashboard and the alarm system, adjust the load orientation between different edge computing nodes to reduce single-point overload and achieve network balance.

[0029] Optionally, identify complex tasks in the edge model and decompose the complex tasks into multiple subtasks, including:

[0030] According to the complexity and characteristics of the tasks, divide the tasks into two types: static tasks and dynamic tasks;

[0031] For the static tasks, establish and deploy task templates and standardized processing procedures on the edge computing nodes. For the dynamic tasks, perform load analysis and dependency analysis to complete the decomposition of the dynamic tasks and obtain multiple subtasks;

[0032] Use a directed acyclic graph model to model the subtasks, clarify the dependency relationships and execution order between the subtasks, and then use a graph partitioning algorithm to divide the directed acyclic graph into multiple subgraphs, and pair the subgraphs with the subtasks.

[0033] Optionally, multiple subtasks are assigned to different edge computing nodes for parallel processing, including:

[0034] Using a reinforcement learning algorithm to formulate a subtask allocation strategy, allocating the subtasks according to the performance of each edge computing node, and establishing a fault tolerance mechanism on the edge computing node to achieve reallocation in case of node failure;

[0035] Determine the task priority according to the urgency, importance, and latency sensitivity of each task, and match the best node type for each task according to the task characteristics;

[0036] Adopt a lightweight transmission protocol and set a differential transmission strategy to compress and transmit only the changed data and newly added data;

[0037] For the task calculation results of each edge computing node, perform correlation analysis and importance evaluation. According to the analysis and evaluation results, use multi-threading technology and parallel computing technology to merge the processing results of each subtask and optimize the merging order of the processing results.

[0038] Optionally, according to the computing process of the edge computing node, perform real-time update and optimization of the edge model, and perform dynamic adjustment of the networking strategy, including:

[0039] Using an edge-terminal collaborative learning mechanism and a federated learning framework to perform cross-node model collaborative training. During the training process, support local fine-tuning of user data on each edge computing node, and only update and train the changed data and newly added data in an incremental training manner;

[0040] Use a dynamic policy table to save the current networking strategy, and perform dynamic adjustment of the networking strategy according to the real-time data analysis results and model output to obtain a new networking strategy, and then use the CI / CD tool chain to gradually deploy the new networking strategy.

[0041] Optionally, introduce encryption technology and access control policies during the data transmission process to protect the data security, and continuously monitor network data and user feedback, including:

[0042] Adopt end-to-end encryption technology and homomorphic encryption algorithm for data encryption, and use blockchain technology for storing encrypted data;

[0043] Allocate access permissions according to user roles, and provide multiple authentication methods for network access in combination with biometric and behavioral characteristics;

[0044] Combined with WiFi probe technology, collect performance metrics on user devices, the core network layer, the edge network layer, and the access network layer for performance monitoring.

[0045] The present invention also provides a 5G intelligent networking system based on edge computing, including:

[0046] An edge deployment module, configured to select user areas and network nodes, deploy multiple edge computing nodes at the selected locations, and design a hierarchical network architecture and a dynamic routing mechanism;

[0047] A data analysis module, configured to collect multi-source data, deploy a lightweight edge model on the edge computing nodes, perform real-time processing and anomaly detection tasks on the multi-source data, and obtain data analysis results;

[0048] A networking policy module, configured to dynamically allocate network resources according to the data analysis results, and perform network switching and load balancing by using the edge model;

[0049] A collaborative computing module, configured to identify complex tasks in the edge model, decompose the complex tasks into multiple subtasks, and allocate the multiple subtasks to different edge computing nodes for parallel processing;

[0050] A collaborative optimization module, configured to perform real-time update and optimization of the edge model according to the computing process of the edge computing nodes, and perform dynamic adjustment of the networking policy;

[0051] A security protection module, configured to introduce encryption technology and access control policies during the data transmission process, perform security protection on the data, and continuously monitor network data and user feedback.

[0052] By providing a 5G intelligent networking method and system based on edge computing, the present invention discloses the following technical effects:

[0053] High network response speed and resource utilization rate: Through a data-driven approach and modular resource configuration, edge computing nodes are flexibly and efficiently deployed near user-dense areas and critical network nodes, reducing data transmission latency. At the same time, by designing a hierarchical network architecture and a dynamic routing mechanism, not only the network response speed and resource utilization rate are improved, but also the 5G network can meet the ever-changing user needs and complex application scenarios.

[0054] Efficient switching and load balancing of networking policies: Through the data analysis results and anomaly detection results of the lightweight edge model, data resource allocation and priority management are performed, and efficient switching and load balancing of networking policies are carried out, enabling balanced cooperation among edge nodes and thus achieving network balance.

[0055] Efficient network processing capabilities and adaptability: Through task decomposition and task partitioning, it is possible to achieve reasonable resource utilization and parallel processing of tasks, reducing network latency and communication overhead; and through edge optimization, multiple edge nodes can perform collaborative training and updates to quickly adapt to changes in various network environments and user needs.

[0056] The following will, through the accompanying drawings and embodiments, further describe in detail the technical solutions of the present invention. Description of the Drawings

[0057] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the accompanying drawings required in the embodiments. Obviously, the accompanying drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other accompanying drawings can also be obtained based on these drawings.

[0058] Figure 1 It is a schematic flowchart of the method provided by the embodiment of the present invention;

[0059] Figure 2 It is a schematic flowchart of the intelligent networking method provided by the embodiment of the present invention;

[0060] Figure 3 It is a schematic diagram of the system architecture provided by the embodiment of the present invention. Detailed Embodiments

[0061] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, rather than all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the scope of protection of the present invention.

[0062] To make the above objects, features, and advantages of the present invention more obvious and understandable, the present invention will be further described in detail below in conjunction with the accompanying drawings and specific embodiments.

[0063] As Figure 1-2 shown, the present invention provides a 5G intelligent networking method based on edge computing, including the following steps:

[0064] 1. Select user areas and network nodes, deploy multiple edge computing nodes at the selected locations, and design a hierarchical network architecture and a dynamic routing mechanism. It includes:

[0065] 1.1 Edge nodes

[0066] Obtain user traffic data and user geographical information through the user device. This can be achieved through data such as connection records of mobile devices and application usage frequencies.

[0067] Use big data analysis technology to analyze the acquired historical data, identify user-dense areas, and based on the user-dense areas, identify key nodes in the network, such as traffic convergence points and areas near data centers. These nodes are usually bottlenecks or key paths in data transmission.

[0068] Deploy multiple edge computing nodes in the user-dense areas and the key nodes.

[0069] Use virtualization technologies (such as Docker, Kubernetes) to dynamically manage and allocate the computing resources of the edge computing nodes, improving resource utilization. And introduce a resource scheduling system to intelligently allocate computing resources according to the real-time load and task requirements of the edge computing nodes, ensuring the efficient operation of the edge nodes.

[0070] 1.2 Network Architecture

[0071] Design a hierarchical network architecture, including:

[0072] Core Network Layer: Perform overall data management and long-distance data transmission, connecting each edge network;

[0073] Edge Network Layer: Located between users and the core network, responsible for local data processing and short-distance transmission, reducing the burden on the core network;

[0074] Access Network Layer: Connect directly to the user device, providing the last-mile access service for the user device.

[0075] Deploy an AI-driven routing algorithm in the hierarchical network architecture for network load analysis, data path management, and edge node adjustment.

[0076] 2. Collect multi-source data, and deploy a lightweight edge model on the edge computing nodes to perform real-time processing and anomaly detection tasks on the multi-source data, obtaining data analysis results. Include:

[0077] 2.1 Data Collection

[0078] User Device Data: Obtain data such as signal strength, location information, connection time, and application usage. The user device can provide this information by installing a lightweight client or using packets in the communication protocol.

[0079] Base Station Data: Collect the load, signal coverage, current active connection count, available bandwidth, etc. of the base stations connected to the users.

[0080] Network device data: Collect traffic statistics data, latency, and packet loss rate of routers and switches that use protocols such as SNMP and NetFlow.

[0081] Integrate the collected data to obtain multi-source data. According to the time dimension and space dimension, use timestamps and geographical location tags to integrate the multi-source data to form a unified data set for synchronous analysis, which can be achieved through timestamps and geographical location tags.

[0082] Use semantic technology to analyze and fuse the integrated data, combine data with similar semantic content, such as combining the bandwidth requirements of different applications with the network device capacity for analysis, to obtain a unified data set.

[0083] 2.2 Edge model

[0084] Select a lightweight deep learning model, such as a convolutional neural network (CNN) or a recurrent neural network (RNN), use the unified data set to train the deep learning model, and set anomaly detection rules (such as threshold detection) in the deep learning model to obtain an edge model.

[0085] Deploy the edge model to the edge computing node to process real-time data and perform anomaly detection to obtain data analysis results.

[0086] 3. Dynamically allocate network resources according to the data analysis results, and use the edge model for network switching and load balancing. Include:

[0087] 3.1 Resource allocation

[0088] According to the data analysis results, monitor the network usage and user needs in real time, identify the usage trends of bandwidth and spectrum, use the edge model to predict changes in network demand, such as burst application traffic or new user connections. Integrate the identification results and prediction results into resource data.

[0089] According to the resource data, take the edge computing node as a unit to dynamically adjust and allocate resources to reduce the burden on the core network.

[0090] 3.2 Network switching

[0091] Use the user's historical movement data to predict the user's movement path and speed to obtain the user's travel data. According to the user's travel data, predict a new site or network node before the user leaves the current base station range for network switching.

[0092] During the switching process, a network protocol that supports seamless switching can be developed to ensure that the user's applications are not affected during the switching process. For the network edge and critical areas, allocate reserve resources in advance for quick takeover.

[0093] 3.3 Load Balancing

[0094] Deploy a real-time monitoring dashboard for displaying the current load status and historical data analysis on the edge computing node, and deploy an alarm system for load alarms to issue early warnings when load overrun or bottlenecks are detected.

[0095] Deploy a load balancing algorithm on the edge computing node. According to the monitoring and alarm results of the monitoring dashboard and the alarm system, adjust the load orientation between different edge computing nodes, reduce single-point overload, realize collaborative work among edge nodes, and assign the tasks of overloaded nodes to nodes with lighter loads to achieve network balance.

[0096] 4. Identify complex tasks in the edge model, decompose complex tasks into multiple subtasks, and assign multiple subtasks to different edge computing nodes for parallel processing. Include:

[0097] 4.1 Task Decomposition

[0098] According to the complexity and characteristics of the tasks, divide the tasks into two types: static tasks and dynamic tasks; static tasks are suitable for processing tasks with high repeatability and clear calculation patterns. Dynamic tasks are divided using real-time analysis data and are suitable for tasks with uncertain calculation patterns.

[0099] For the static tasks, establish and deploy task templates and standardized processing procedures on the edge computing node. For dynamic tasks, perform load analysis and dependency analysis to complete the decomposition of the dynamic tasks and obtain multiple subtasks.

[0100] Use a directed acyclic graph model to model the subtasks, clarify the dependency relationships and execution order between each subtask, and then use graph partitioning algorithms (such as METIS, SCOTCH) to divide the directed acyclic graph into multiple subgraphs, and pair the subgraphs with the subtasks so that each subgraph corresponds to a subtask that can be independently executed on an edge node.

[0101] 4.2 Subtask Allocation and Parallel Processing

[0102] Use a reinforcement learning algorithm to formulate a subtask allocation strategy, allocate the subtasks according to the performance of each edge computing node, and establish a fault tolerance mechanism on the edge computing node to achieve reallocation in case of node failure and ensure the continuity of tasks.

[0103] Determine the task priority according to the urgency, importance, and latency sensitivity of each task, and match the best node type for each task according to the task characteristics (such as CPU-intensive, GPU-intensive, I / O-intensive).

[0104] Adopt lightweight transport protocols, such as efficient transport protocols like gRPC and QUIC, and set differential transmission strategies to only compress and transmit changed data and newly added data, reducing the data volume and bandwidth occupancy.

[0105] For the task calculation results of each of the edge computing nodes, conduct correlation analysis and importance evaluation. According to the analysis and evaluation results, utilize multi-threading technology and parallel computing technology to accelerate the merging of the processing results of each sub-task, and optimize the merging order of the processing results.

[0106] 5. According to the calculation process of the edge computing nodes, conduct real-time update and optimization of the edge model, and conduct dynamic adjustment of the networking strategy. It includes:

[0107] Utilize the edge-terminal collaborative learning mechanism and the federated learning framework to conduct cross-node model collaborative training. During the training process, support local fine-tuning of user data on each of the edge computing nodes, and through incremental training, only update and train changed data and newly added data to minimize the burden of model update, while ensuring the ability to quickly adapt to environmental changes.

[0108] Use a dynamic policy table to save the current networking strategy, and conduct dynamic adjustment of the networking strategy based on real-time data analysis results and model outputs to obtain a new networking strategy, and then utilize the CI / CD tool chain to gradually deploy the new networking strategy. The gradual deployment reduces the uncertain impact of the new strategy on the overall network by deploying in stages or regions.

[0109] 6. Introduce encryption technology and access control strategies during the data transmission process to conduct data security protection, and continuously monitor network data and user feedback. It includes:

[0110] Adopt end-to-end encryption technology and homomorphic encryption algorithms for data encryption, and use blockchain technology for the storage of encrypted data.

[0111] Allocate access permissions according to user roles, and combine biometric and behavioral characteristics to provide multiple authentication methods for network access.

[0112] Combine WiFi probe technology to collect performance metrics on user devices, the core network layer, the edge network layer, and the access network layer for performance monitoring.

[0113] As Figure 3 shown, the present invention also provides a 5G intelligent networking system based on edge computing, including:

[0114] An edge deployment module for selecting user areas and network nodes, deploying multiple edge computing nodes at the selected locations, and designing a hierarchical network architecture and a dynamic routing mechanism.

[0115] A data analysis module for collecting multi-source data, deploying lightweight edge models on the edge computing nodes, performing real-time processing and anomaly detection tasks on the multi-source data, and obtaining data analysis results.

[0116] A networking policy module for dynamically allocating network resources according to the data analysis results, and using the edge model for network switching and load balancing.

[0117] A collaborative computing module for identifying complex tasks in the edge model, decomposing the complex tasks into multiple subtasks, and allocating the multiple subtasks to different edge computing nodes for parallel processing.

[0118] A collaborative optimization module for performing real-time update and optimization of the edge model according to the computing process of the edge computing nodes, and dynamically adjusting the networking policy.

[0119] A security protection module for introducing encryption technologies and access control policies during data transmission, performing data security protection, and continuously monitoring network data and user feedback.

[0120] Therefore, by providing a 5G intelligent networking method and system based on edge computing, the present invention can efficiently utilize network resources, perform intelligent networking and network switching, achieve load balancing, greatly improve the network response speed, processing capacity and adaptability, and enhance the user experience.

[0121] The various embodiments in this specification are described in a progressive manner. Each embodiment focuses on the differences from other embodiments. The same or similar parts among the various embodiments can be referred to each other.

[0122] Specific examples are used in this article to elaborate on the principles and implementation manners of the present invention. The descriptions of the above embodiments are only used to help understand the method and its core idea of the present invention; at the same time, for those of ordinary skill in the art, according to the idea of the present invention, there will be changes in the specific implementation manners and application scopes. In summary, the content of this specification should not be construed as a limitation to the present invention.

Claims

1. A 5G intelligent networking method based on edge computing, characterized in that, It includes the following steps: Select user-dense areas and key network nodes based on user historical data, deploy multiple edge computing nodes at the selected locations, and design a hierarchical network architecture including a core network layer, an edge network layer, and an access network layer, as well as a dynamic routing mechanism; Collect multi-source data including user device data, base station data, and network device data, and deploy lightweight edge models on the edge computing nodes to perform real-time processing and anomaly detection tasks on the multi-source data, obtaining data analysis results; Dynamically allocate network resources according to the data analysis results, and use the edge model to predict new sites or network nodes based on the user travel data obtained from the user movement path and movement speed predicted by the user historical movement data for network switching, as well as perform load balancing based on the real-time monitoring dashboard and load balancing algorithm; Identify complex tasks in the edge model, decompose the complex tasks into multiple subtasks, and use reinforcement learning algorithms and fault tolerance mechanisms to allocate the multiple subtasks to different edge computing nodes for parallel processing; Use the edge-terminal collaborative learning mechanism and the federated learning framework to perform cross-node model collaborative training. During the training process, support local fine-tuning of user data on each edge computing node, and through incremental training, only update and train the changed data and newly added data; Use a dynamic policy table to save the current networking policy, and perform dynamic adjustment of the networking policy according to the real-time data analysis results and model outputs to obtain a new networking policy, and then use the CI / CD toolchain to gradually deploy the new networking policy; Introduce encryption technology and access control policies during the data transmission process to protect data security, and continuously monitor network data and user feedback.

2. The 5G intelligent networking method based on edge computing according to claim 1, wherein, Select user areas and network nodes, and deploy multiple edge computing nodes at the selected locations, including: Obtain user traffic data and user geographical information through user devices, analyze the obtained historical data using big data analysis technology, identify user-dense areas, and based on the user-dense areas, identify key nodes in the network; Deploy multiple edge computing nodes in the user-dense areas and the key nodes; Use virtualization technology to dynamically manage the computing resources of the edge computing nodes, and introduce a resource scheduling system to intelligently allocate computing resources according to the real-time load and task requirements of the edge computing nodes; Design a hierarchical network architecture and deploy an AI-driven routing algorithm in the hierarchical network architecture to perform network load analysis, data path management, and edge node adjustment; Among them, the hierarchical network architecture includes a core network layer for overall data management and long-distance data transmission, an edge network layer for local data processing and short-distance transmission, and an access network layer for providing user device access services.

3. The 5G intelligent networking method based on edge computing according to claim 2, characterized in that, Collect multi-source data, and deploy lightweight edge models on the edge computing nodes to perform real-time processing and anomaly detection tasks on the multi-source data, obtaining data analysis results, including: Obtain user device data, base station data, and network device data to obtain multi-source data; Integrate the multi-source data according to the time dimension and space dimension, using timestamps and geographical location tags, and then use semantic technology to analyze and fuse the integrated data to obtain a unified data set; Select a lightweight deep learning model, use the unified data set to train the deep learning model, and set anomaly detection rules in the deep learning model to obtain an edge model; Deploy the edge model to the edge computing node to process real-time data and perform anomaly detection to obtain data analysis results.

4. A 5G intelligent networking method based on edge computing according to claim 3, wherein The user equipment data includes signal strength, location information, connection time, and application usage. The base station data includes base station load, signal coverage, current active connection count, and available bandwidth. The network equipment data includes router traffic statistics, switch traffic statistics, latency, and packet loss rate.

5. A 5G intelligent networking method based on edge computing according to claim 4, characterized in that, Dynamically allocate network resources according to the data analysis results, and use the edge model for network switching and load balancing, including: According to the data analysis results, monitor the network usage and user requirements in real time, identify the usage trends of bandwidth and spectrum, and use the edge model to predict changes in network requirements, and integrate the identification results and prediction results into resource data; According to the resource data, perform dynamic adjustment and allocation of resources in units of the edge computing node; Predict the user's movement path and movement speed using the user's historical movement data to obtain user travel data. According to the user travel data, predict a new site or network node before the user leaves the current base station range for network switching; Deploy a real-time monitoring dashboard for displaying the current load status and historical data analysis on the edge computing node, and deploy an alarm system for load alerts; Deploy a load balancing algorithm on the edge computing node. According to the monitoring and alarm results of the monitoring dashboard and the alarm system, adjust the load orientation between different edge computing nodes to reduce single-point overload and achieve network balance.

6. A 5G intelligent networking method based on edge computing according to claim 5, characterized in that Identify complex tasks in the edge model and decompose the complex tasks into multiple subtasks, including: According to the complexity and characteristics of the tasks, divide the tasks into two types: static tasks and dynamic tasks; For the static tasks, establish and deploy task templates and standardized processing procedures on the edge computing node. For the dynamic tasks, perform load analysis and dependency analysis to complete the decomposition of the dynamic tasks and obtain multiple subtasks; Use a directed acyclic graph model to model the subtasks, clarify the dependency relationships and execution orders between the subtasks, and then use a graph partitioning algorithm to divide the directed acyclic graph into multiple subgraphs and pair the subgraphs with the subtasks.

7. A 5G intelligent networking method based on edge computing according to claim 6, wherein Allocate multiple subtasks to different edge computing nodes for parallel processing, including: Use a reinforcement learning algorithm to formulate a subtask allocation strategy, allocate the subtasks according to the performance of each edge computing node, and establish a fault tolerance mechanism on the edge computing node to achieve reallocation in case of node failure; Determine the task priority according to the urgency, importance, and latency sensitivity of each task, and match the best node type for each task according to the task characteristics; Adopt a lightweight transmission protocol and set a differential transmission strategy to only compress and transmit changed data and newly added data; For the task calculation results of each of the edge computing nodes, perform correlation analysis and importance assessment. According to the analysis and assessment results, use multi-threading technology and parallel computing technology to merge the processing results of each of the subtasks, and optimize the merging order of the processing results.

8. A 5G intelligent networking method based on edge computing according to claim 7, characterized in that Introduce encryption technology and access control policies during the data transmission process to protect data security, and continuously monitor network data and user feedback, including: Adopt end-to-end encryption technology and homomorphic encryption algorithms for data encryption, and use blockchain technology for storing encrypted data; Allocate access permissions according to user roles, and combine biometric and behavioral characteristics to provide multiple authentication methods for network access; Combine WiFi probe technology to collect performance metrics on user devices, the core network layer, the edge network layer, and the access network layer for performance monitoring.

9. A 5G intelligent networking system based on edge computing, characterized in that, Include: An edge deployment module, which is used to select user-dense areas and key network nodes based on user historical data, deploy multiple edge computing nodes at the selected locations, and design a hierarchical network architecture including a core network layer, an edge network layer, and an access network layer, as well as a dynamic routing mechanism; A data analysis module, which is used to collect multi-source data including user device data, base station data, and network device data, and deploy a lightweight edge model on the edge computing nodes to perform real-time processing and anomaly detection tasks on the multi-source data, and obtain data analysis results; A network configuration strategy module, which is used to dynamically allocate network resources according to the data analysis results, and use the edge model to predict new sites or network nodes for network switching based on the user travel data obtained from the predicted user movement paths and movement speeds based on user historical movement data, and perform load balancing based on a real-time monitoring dashboard and a load balancing algorithm; A collaborative computing module, which is used to identify complex tasks in the edge model, decompose complex tasks into multiple subtasks, and use reinforcement learning algorithms and fault tolerance mechanisms to allocate multiple subtasks to different edge computing nodes for parallel processing; A collaborative optimization module, which is used to use an edge-terminal collaborative learning mechanism and a federated learning framework to perform cross-node model collaborative training. During the training process, support local fine-tuning of user data on each of the edge computing nodes, and only update and train changed data and newly added data in an incremental training manner; Use a dynamic policy table to save the current network configuration strategy, and dynamically adjust the network configuration strategy according to real-time data analysis results and model outputs to obtain a new network configuration strategy, and then use a CI / CD toolchain to gradually deploy the new network configuration strategy; A security protection module, which is used to introduce encryption technology and access control policies during the data transmission process to protect data security, and continuously monitor network data and user feedback.

Images