Intelligent routing optimization method and system in high-speed optical network

Abstract

The invention relates to the technical field of data transmission control, and discloses an intelligent routing optimization method and system in a high-speed optical network, and the method comprises the steps: obtaining the structure and performance information of a target area network, constructing a network simulation model, setting a network link target according to a to-be-executed network link task, and carrying out the execution of the network link target; the performance redundancy requirement of a network model is adjusted by predicting and analyzing external task interference and future task change, a network link target and the redundancy requirement are simulated based on the model, an initial network topology structure is obtained, and network link processing is executed; according to the method, task status data are continuously collected, the redundancy requirement is synchronously corrected according to the data, the initial topology is optimized, the real-time optimized synchronous network topology is obtained, and the problem that in the prior art, a static routing and topology optimization method often cannot cope with frequently-changing network requirements is solved.

Description

Technical Field

[0001] The present invention relates to the technical field of data transmission control, and in particular to an intelligent routing optimization method and system in a high-speed optical network. Background Art

[0002] With the rapid development of information technology and communication technology, the demand for data transmission continues to increase, especially in the fields of cloud computing, big data, and the Internet of Things, making high-speed optical networks a core infrastructure. As the main force for carrying large-scale data transmission, high-speed optical networks face the problems of complexity and dynamic changes in network topology while meeting extremely high bandwidth requirements, low latency, and high reliability.

[0003] In high-speed optical networks, the working status of links and nodes is highly dynamic. Due to a large number of concurrent tasks and the real-time load of the network, traditional static routing and topology optimization methods often cannot cope with frequently changing network requirements. Summary of the invention

[0004] The purpose of the present invention is to provide an intelligent routing optimization method and system in a high-speed optical network, aiming to solve the problem that static routing and topology optimization methods in the prior art are often unable to cope with frequently changing network requirements.

[0005] The present invention is implemented in this way. In a first aspect, the present invention provides an intelligent routing optimization method in a high-speed optical network, comprising: Acquiring network structure information and network performance information of a target area network, and performing digital modeling processing on the target area network according to the network structure information and the network performance information to obtain a network simulation model of the target area network; Acquire each network link task to be executed by the target area network at the current moment, and set the network link target for the network simulation model according to each network link task; Performing a predictive analysis of external task interference and future task changes on the network simulation model, and setting performance redundancy requirements for the network simulation model based on the analysis results; Performing network link simulation on the network link target and the performance redundancy requirement based on the network simulation model to obtain an initial link topology structure of the target area network, and performing network link processing on the target area network according to the initial link topology structure; The target area network continuously collects the real-time status of the mission to obtain the real-time characteristics of the mission, and synchronously corrects the performance redundancy requirements of the network simulation model according to the real-time characteristics of the mission, so as to drive the network simulation model to perform real-time optimization on the initial link topology structure, so as to obtain the synchronous link topology structure of the target area network and perform network link processing on the target area network.

[0006] In a second aspect, the present invention provides an intelligent routing optimization system in a high-speed optical network, which is used to implement an intelligent routing optimization method in a high-speed optical network as described in any one of the first aspects, including: A network simulation module, used to obtain network structure information and network performance information of a target area network, and perform digital modeling processing on the target area network according to the network structure information and the network performance information to obtain a network simulation model of the target area network; A target setting module, used for obtaining each network link task to be executed by the target area network at the current moment, and setting the network link target for the network simulation model according to each network link task; A redundancy setting module, used to perform a predictive analysis of external task interference and future task changes on the network simulation model, and to set performance redundancy requirements for the network simulation model according to the analysis results; A link simulation module, configured to perform network link simulation on the network link target and the performance redundancy requirement based on the network simulation model to obtain an initial link topology structure of the target area network, and perform network link processing on the target area network according to the initial link topology structure; A real-time optimization module is used to continuously collect the real-time status of the task status of the target area network to obtain the real-time characteristics of the task, and to synchronously correct the performance redundancy requirements of the network simulation model according to the real-time characteristics of the task, so as to drive the network simulation model to perform real-time optimization on the initial link topology structure, so as to obtain the synchronous link topology structure of the target area network and perform network link processing on the target area network.

[0007] The present invention provides an intelligent routing optimization method in a high-speed optical network, which has the following beneficial effects: The present invention acquires the structure and performance information of the target area network to construct a network simulation model, sets the network link target according to the network link task to be executed, adjusts the performance redundancy requirement of the network model by predicting and analyzing the external task interference and future task changes, simulates the network link target and redundancy requirement based on the model, obtains the initial network topology structure and executes the network link processing, continuously collects the task status data and synchronously corrects the redundancy requirement according to the data to optimize the initial topology, obtains the real-time optimized synchronous network topology, and solves the problem that the static routing and topology optimization methods in the prior art are often unable to cope with the frequently changing network requirements. BRIEF DESCRIPTION OF THE DRAWINGS

[0008] Figure 1 It is a schematic diagram of the steps of an intelligent routing optimization method in a high-speed optical network provided by an embodiment of the present invention; Figure 2 The invention is a schematic diagram of the structure of an intelligent routing optimization system in a high-speed optical network provided by an embodiment of the invention. DETAILED DESCRIPTION

[0009] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0010] The implementation of the present invention is described in detail below in conjunction with specific embodiments.

[0011] Reference Figure 1 , Figure 2 As shown, a preferred embodiment of the present invention is provided.

[0012] In a first aspect, the present invention provides an intelligent routing optimization method in a high-speed optical network, comprising: S1: obtaining network structure information and network performance information of a target area network, and performing digital modeling processing on the target area network according to the network structure information and the network performance information to obtain a network simulation model of the target area network; S2: Acquire each network link task to be executed by the target area network at the current moment, and set the network link target for the network simulation model according to each network link task; S3: performing a predictive analysis of external task interference and future task changes on the network simulation model, and setting performance redundancy requirements for the network simulation model according to the analysis results; S4: performing network link simulation on the network link target and the performance redundancy requirement based on the network simulation model to obtain an initial link topology structure of the target area network, and performing network link processing on the target area network according to the initial link topology structure; S5: Continuously collect the real-time status of the task on the target area network to obtain the real-time characteristics of the task, and synchronously modify the performance redundancy requirements of the network simulation model according to the real-time characteristics of the task, so as to drive the network simulation model to perform real-time optimization on the initial link topology structure, so as to obtain the synchronous link topology structure of the target area network and perform network link processing on the target area network.

[0013] Specifically, in step S1 of the embodiment provided by the present invention, information of all network nodes in the target area network is collected. These nodes may include switches, routers, servers, etc., which are basic components of network communication. Links are physical or virtual paths connecting these nodes, usually involving optical fiber links, cables, wireless connections, etc. It is necessary to obtain information such as link bandwidth, transmission rate, delay, signal strength, etc., and identify the overall topology of the target area network through node and link information. The network topology can be tree-like, ring-like, mesh-like or star-like, etc., which describes the connection mode between nodes.

[0014] More specifically, the performance indicators of each node (such as processing power, load, response time, bandwidth, etc.) need to be collected. The performance of network nodes directly affects the data processing speed and the network's carrying capacity. The performance of the link usually includes latency, bandwidth, transmission quality (such as packet loss rate, bit error rate), etc., which will affect the efficiency of data transmission. The network load condition, bandwidth usage, link load balancing, etc. are also performance data that need to be obtained, especially in high-traffic networks, where load balance and link utilization are crucial to performance.

[0015] More specifically, a digital network model is constructed based on the collected network structure information and performance information. This model not only reflects the physical topology of the network, but also can simulate the performance of the network, including node performance, link latency, bandwidth utilization, etc. A digital model in the form of a topology can be used to represent the network topology. Nodes represent network devices, and links represent communication paths between devices. Relevant performance data (such as bandwidth, latency, load, etc.) are attached to each link and each node. For dynamic characteristics (such as changes in network load, task changes, etc.), a timing model or event-driven model can be introduced to simulate the performance of the network over time.

[0016] More specifically, in order to simulate the real performance of the network, it is usually necessary to use network simulation tools (such as NS3, OMNeT++, Mininet, etc.) to build and test network simulation models. Through these simulation tools, different network tasks, traffic patterns, equipment failures, network attacks and other scenarios can be simulated. After the network model is built, preliminary verification is required to check the accuracy of the model by comparing the results of the network model with the behavior of the actual network. If deviations are found, the parameters in the model (such as node performance, link bandwidth, etc.) need to be further adjusted. In the subsequent network optimization process, the network model can be dynamically adjusted. By introducing algorithms (such as machine learning, genetic algorithms, etc.) to optimize the parameters in the model, the prediction accuracy of the model and the overall performance of the network can be improved.

[0017] It is understandable that by building a digital network model based on real data, the topology, performance and behavior of the network can be accurately simulated. This simulation model can help researchers and engineers better understand the operating characteristics of the network and predict the performance of the network under different conditions. Through the digital modeling of network structure and performance information, it is easier to apply optimization algorithms to improve network performance. For example, optimizing route selection, load balancing, bandwidth allocation, etc. to ensure efficient operation of the network can also help predict future changes in network demand and provide a basis for capacity planning and resource scheduling. Combined with task demand prediction, redundancy design and performance assurance can be carried out in advance during the network structure design stage.

[0018] Specifically, in step S2 of the embodiment provided by the present invention, various network tasks to be executed in the target area network at the current moment are identified. These tasks may be data transmission requests, real-time video streaming, large-scale file uploads, network traffic scheduling, etc. Each task will have different requirements, for example: Bandwidth requirements: Different tasks may require different bandwidths. For example, high-definition video streaming requires higher bandwidth, while general data transmission tasks require lower bandwidth. Delay requirements: Some real-time tasks (such as voice communication, online games) are very sensitive to delay and require as low delay as possible, while other tasks may be insensitive to delay. Reliability requirements: For example, for important business data transmission, high reliability is required, while other tasks may tolerate a certain amount of packet loss or error.

[0019] More specifically, by classifying task types, tasks can be divided into different priorities and reasonably scheduled according to the characteristics of each task (such as real-time, bandwidth requirements, data volume, etc.). Task scheduling can be processed using methods such as time slice rotation and priority queues. According to the characteristics of each task, its requirements are mapped to the network simulation model. This mapping process mainly includes: Bandwidth allocation: According to the bandwidth requirements of the task, sufficient bandwidth resources are allocated to the task. For example, high-traffic tasks will obtain higher bandwidth, while the bandwidth of low-traffic tasks can be appropriately limited. Path selection: Select a suitable transmission path for each task. The path selection needs to consider the network topology, link bandwidth, delay, Factors such as load. If the task is sensitive to delay, a shorter and more direct path is selected; if the task is not sensitive to delay, a path with lower load can be selected to balance resources. For tasks with high reliability requirements, redundant paths can be set to ensure that the task can be completed smoothly when the main path fails. Set priority and quality of service (QoS): Set different priorities for different tasks and specify corresponding quality of service requirements (QoS). For example, real-time tasks (such as video conferencing and voice communication) can be set with higher priorities and ensure lower delays and packet loss rates. Non-real-time tasks (such as file downloads and big data analysis) can be set with lower priorities and accept certain delays or packet loss.

[0020] More specifically, after the network simulation model is established, the routing, bandwidth allocation, topology and other parameters in the network simulation model are adjusted based on the link target of each task. This process can be carried out in the following ways: According to the task requirements and the current network conditions (such as link load, node available bandwidth, etc.), the network topology is dynamically adjusted so that the task can be transmitted on the most appropriate path. If the load of some links or nodes is too high, the system can dynamically adjust the routing path of the task to ensure network load balance and avoid overload. According to the setting of the above network link targets, the network simulation model is used for simulation analysis to verify whether the various requirements of the task can be met. If the network resources (bandwidth, nodes, links) are not enough to meet the requirements of certain tasks, they can be adjusted through optimization algorithms (such as shortest path algorithm, traffic scheduling algorithm). At the same time, network performance evaluation can also be performed to analyze link delays, bandwidth occupancy, packet loss rate, etc. to ensure that the allocation and scheduling of network resources are reasonable.

[0021] More specifically, by accurately analyzing the needs of each task and intelligently allocating bandwidth, routing, and other resources, we can effectively avoid resource waste and maximize network efficiency. Tasks will be reasonably scheduled according to their priority and needs to ensure that high-priority tasks (such as real-time tasks) obtain sufficient network resources.

[0022] Specifically, in step S3 of the embodiment provided by the present invention, load information of external tasks is identified and collected. These tasks come from outside the network (such as other network users, external application systems, devices in other geographical locations, etc.). The traffic and load characteristics of these tasks are crucial to the performance of the network. External interference may manifest as burst traffic (for example, DDoS attacks, burst data traffic) or other uncertain events (such as network equipment failure, link failure, etc.). It is necessary to predict the impact of these interferences on the network and take appropriate preventive measures. Through historical data and trend analysis, the future load changes or traffic fluctuations of external tasks are predicted. This can be accomplished through time series analysis and machine learning models (such as regression analysis, neural networks, etc.).

[0023] More specifically, the tasks to be executed in the target area network in the future are obtained, including parameters such as the task start time, duration, bandwidth requirements, and delay requirements. Time series prediction or machine learning methods (such as support vector machines, random forests, etc.) are used to model the dynamic changes of future tasks, and the demand and load of network resources caused by task changes (such as new tasks, changes in task priorities, etc.) are analyzed. This helps to evaluate the impact of task changes on network structure, link selection, bandwidth allocation, etc. in advance.

[0024] More specifically, when external task interference or future task changes are predicted, performance redundancy requirements need to be set. Performance redundancy means reserving sufficient network resources (such as additional bandwidth, backup links, etc.) to ensure network stability in order to cope with situations such as increased network load, link failures, and burst traffic. Redundancy requirements are dynamically adjusted based on predicted external interference and task changes. For example, when the network predicts a high traffic impact, the bandwidth redundancy of the link can be increased or backup nodes can be added. When equipment failure or link loss is predicted, redundant paths can be set to ensure that data transmission is not affected.

[0025] More specifically, by finely setting different levels of redundancy, it is possible to avoid wasting redundant resources while ensuring network stability. For example, a higher redundancy level can be set for the traffic of high-priority tasks, while a lower redundancy level can be set for low-priority tasks.

[0026] More specifically, the actual performance of external interference and task changes in the network is continuously monitored, and the monitoring data is fed back to the network simulation model to further optimize the setting of performance redundancy. This enables the network to make real-time adjustments based on dynamic changes in actual operation. Based on the feedback information, the model can continuously optimize the setting of redundancy requirements. For example, if the actual load changes of external tasks are more drastic than expected, the network can further adjust redundant resources to ensure that network performance is not affected.

[0027] It is understandable that by predicting external task interference and future task changes and setting performance redundancy requirements in advance, the network can cope with sudden traffic, task changes, and link or node failures, which makes the network more stable and more fault-tolerant in the face of an uncertain external environment. Redundant design can effectively prevent performance degradation due to changes in network load or equipment failure, ensuring that the network continues to provide stable services. By reasonably setting redundancy requirements, the network can accurately allocate resources and avoid waste of resources due to excessive redundancy. Dynamic adjustment of redundancy levels can ensure that network resources are optimally configured in different scenarios, reducing unnecessary waste of bandwidth and computing resources. At the same time, when traffic shocks or task changes are predicted, resources can be quickly scheduled to improve network utilization.

[0028] More specifically, through the dynamic setting of redundancy requirements, the transmission quality of high-priority tasks can be guaranteed not to be affected. Even when the network load increases or emergencies occur, important tasks can still guarantee the required bandwidth, low latency and other performance requirements. By analyzing future task changes, the most appropriate resource quota can be set for the task to avoid network resource congestion, thereby improving network response speed and transmission stability. The network simulation model can automatically adjust the redundancy requirements according to the prediction of external interference and task changes, thereby achieving adaptive adjustment. This adaptive capability enables the network to respond quickly to changes in real-time needs, reduce human intervention, and improve operational efficiency.

[0029] Specifically, in step S4 of the embodiment provided by the present invention, the network connection target is clarified, including: task requirements: such as delay, bandwidth, throughput, priority, etc. These task requirements may come from different business types (such as video streaming, voice communication, large data transmission, etc.), and each task has different performance requirements, network connection targets: for example, a stable communication path needs to be established between certain nodes or devices, or certain services require a dedicated link with high bandwidth and low latency, and performance redundancy requirements are set: according to the priority of the task and the requirements of network performance, a redundancy strategy is set. For example, some key links may need to provide redundant channels to ensure normal communication in the event of a link failure, and other links may not require redundancy.

[0030] More specifically, according to the network link goals and performance redundancy requirements, network link simulation is performed in the network simulation model. This step will use data such as network topology, link bandwidth, latency, and redundancy requirements for simulation and evaluate different link schemes. According to task requirements, the simulation system will select appropriate links to meet network connection goals and choose whether to add backup links based on performance redundancy requirements. For links that require redundancy, the simulation will introduce backup links. During the simulation, different link failures and task changes can be simulated to verify whether the redundant links can successfully take over the traffic and maintain the stability of the network. During the simulation, the traffic scheduling scheme must be dynamically adjusted according to the timing characteristics and traffic requirements of the task to ensure that the traffic of each task in the network is transmitted smoothly and meets its performance requirements.

[0031] More specifically, through the above simulation, the system generates a preliminary link topology of the network. This topology includes all nodes and links in the network, as well as network paths allocated according to task requirements and redundancy requirements. When generating the initial topology, the simulation process will automatically optimize to ensure that the topology not only meets the network link goals, but also maximizes resource utilization and avoids bottlenecks. For example, select the link with the most appropriate bandwidth to transmit task traffic to avoid network congestion and ensure that redundant paths are taken into account in the topology. Even if a node or link failure occurs, the network can still maintain stable communication through redundant links.

[0032] More specifically, according to the generated initial link topology, further network link processing is performed. This involves specific link configuration, routing settings and bandwidth allocation. At this stage, further algorithms (such as the shortest path algorithm, maximum flow algorithm, load balancing algorithm, etc.) are used to optimize network links to ensure balanced distribution of network resources (especially bandwidth) to improve overall network utilization and service quality. For parts with redundant links, when the main link fails, the backup link will automatically take over the traffic, thereby ensuring uninterrupted network services.

[0033] More specifically, simulation tests are conducted on the network link topology structure to evaluate the network's performance under different task loads, traffic changes, link failures, etc., to test whether the network can meet the preset delay requirements, bandwidth requirements, task priorities, etc., to simulate link or node failure scenarios, and to check the switching of redundant links to ensure that the network can quickly restore services through redundant links or other recovery mechanisms when a failure occurs, thereby ensuring the high availability of the network.

[0034] It can be understood that through simulation based on network link targets and performance redundancy requirements, the generated network link topology can more accurately reflect the actual needs of the target area network, including task load, link bandwidth, delay and other performance requirements, which provides a reliable foundation for subsequent network optimization and resource scheduling. Through dynamic simulation and optimization, the system can efficiently configure network resources (such as bandwidth, links, etc.), avoid resource waste and link congestion problems, and improve network resource utilization. Through the setting of redundant links, it is ensured that the network can cope with load changes in different scenarios without causing network performance degradation due to insufficient resources. The generated topology can automatically adapt to link or node failures. The activation of redundant links ensures that the network can recover quickly when a failure occurs, thereby improving the network's fault tolerance and reliability. The network can better cope with external interference and task changes, ensure that critical tasks are not affected, and improve the stability of the network.

[0035] More specifically, through precise link selection and load balancing, it is possible to ensure that network services meet the quality of service (QoS) requirements of different tasks, such as latency, bandwidth, packet loss rate, etc. This is especially important for high-priority tasks. It can guarantee dedicated links with high bandwidth and low latency. The network can be dynamically adjusted according to the different requirements of the tasks to ensure that each task can obtain appropriate network resources. The initial link topology and the results of network link processing not only meet current needs, but also support the increase in network capacity and new task requirements through dynamic expansion. For example, new links or nodes are added to cope with the increase in future task loads. By simulating future task changes and traffic changes, the network can plan capacity in advance according to demand to ensure that the network can still maintain efficient operation during expansion. The initial link topology is generated and processed through the network simulation model, which can speed up the network optimization and deployment cycle and reduce the problems that may arise during trial and error and actual deployment. This provides technical support for the rapid deployment and continuous optimization of the network.

[0036] Specifically, in step S5 of the embodiment provided by the present invention, the running status of various tasks in the target area network is continuously monitored, and the current status characteristics of the tasks include: task traffic: bandwidth requirements, data packet size, transmission frequency, etc. of the current task, delay and packet loss rate: delay requirements and packet loss of the task, task priority and quality of service (QoS): different tasks may have different priorities and quality of service requirements, such as high-priority tasks may require low delay and high bandwidth, dynamic changes in tasks: the traffic, priority and delay requirements of the task may be dynamically adjusted over time, and traffic monitoring tools (such as SNMP, NetFlow, sFlow, etc.) and performance monitoring tools (such as ping, traceroute, etc.) are used to collect task traffic, delay, etc. in the network in real time and transmit them to the network control center.

[0037] More specifically, the task status data collected in real time is analyzed to extract the task's changing trends, performance bottlenecks, traffic patterns, etc. The specific analysis contents include: Traffic change analysis: identifying the fluctuation trend of task traffic, such as sudden increase or decrease in traffic, delay and packet loss analysis: analyzing the delay changes and packet loss of the task, and evaluating whether there are network bottlenecks, service quality demand changes: tracking the service quality requirements of the task to ensure that the QoS requirements of the task are always met.

[0038] More specifically, based on the actual characteristics of the tasks obtained through analysis, the redundancy requirements in the network simulation model are revised. This is mainly reflected in the need to add redundant links to ensure network reliability if the bandwidth demand of the task increases suddenly, or the network connection of some tasks is unstable. If the delay requirements of the task become more stringent, or the bandwidth requirements change, the network simulation model needs to dynamically adjust the number of redundant links and bandwidth allocation strategies to ensure that the task requirements are met. According to the real-time changes in task traffic, the activation strategy and switching strategy of the redundant links are adjusted. For example, when a network link fails, the switching delay of the redundant links is optimized to ensure that the task is not affected.

[0039] More specifically, based on the actual characteristics of the mission and the revised performance redundancy requirements, the network simulation model is driven to optimize the initial link topology. The optimization process includes: dynamically adjusting link selection according to mission requirements, selecting links with optimal bandwidth, delay and redundancy requirements, dynamically adjusting traffic distribution plans according to the real-time load of the mission to ensure load balancing of each link to avoid bottlenecks caused by overloaded links, adding new nodes or links to the network topology to adapt to changing mission requirements, or migrating traffic from overloaded links to less busy links to avoid link congestion.

[0040] More specifically, a new synchronization link topology is generated through the optimization of the network simulation model. This topology reflects the real-time changes of the current network status and mission requirements, which may be different from the initial topology and has better adaptability and flexibility. The optimized synchronization link topology will be applied to the target area network in real time to ensure that the network's connectivity, bandwidth, latency and other performance meet the mission requirements and meet the requirements of redundant links and fault tolerance.

[0041] More specifically, based on the new synchronous link topology, network link processing is performed, including: link configuration and routing selection: according to the new topology, configure the links, routers and switches in the network to ensure that the network path meets the performance requirements of the task, traffic scheduling and load balancing: implement traffic scheduling in the network to ensure that each task runs on the optimal link to avoid network bottlenecks, redundant link activation: according to the changes in the network topology, activate the redundant links in a timely manner to ensure that the task can maintain normal transmission in the event of link failure or congestion.

[0042] It is understandable that through real-time collection and analysis of the current status of tasks, the network can adjust according to the dynamic changes of tasks. For example, when the traffic of a task surges, the system can automatically add redundant links or adjust the routing path to ensure that the task can be completed smoothly. The network simulation model optimizes the topology structure in real time to achieve efficient use of network resources. Through load balancing and link optimization, it avoids resource waste and overloaded links, ensures that the bandwidth of each link is reasonably allocated, and improves the performance and efficiency of the overall network. Through dynamic optimization of redundant links and fault-tolerant mechanisms, the network can quickly respond to link failures or congestion, improve network reliability, and even when a link fails or a performance bottleneck occurs, redundant links can take over the task in time to ensure that the task is not interrupted.

[0043] More specifically, by dynamically adjusting task requirements, the Quality of Service (QoS) requirements of different tasks are ensured to be met in real time. For example, priority bandwidth and lower latency are provided to high-priority tasks to prevent low-priority tasks from affecting the transmission performance of high-priority tasks. Real-time task status collection and synchronous correction of network simulation models enable the network to quickly respond to factors such as network topology changes, link failures, and traffic fluctuations, ensuring that the network is always in the optimal state. By continuously collecting and optimizing the actual characteristics of tasks, the network can better adapt to uncertain changes in task requirements in the future. Whether it is task traffic, latency requirements or redundancy requirements, the network can be corrected and optimized in the shortest time.

[0044] More specifically, during the network topology optimization process, the system can evaluate and plan network expansion in real time to ensure that when the task load grows in the future, the network can expand smoothly without affecting existing services. At the same time, the maintainability of the network is also enhanced. Through real-time monitoring and feedback mechanisms, potential problems can be discovered and resolved in a timely manner.

[0045] The present invention provides an intelligent routing optimization method in a high-speed optical network, which has the following beneficial effects: The present invention acquires the structure and performance information of the target area network to construct a network simulation model, sets the network link target according to the network link task to be executed, adjusts the performance redundancy requirement of the network model by predicting and analyzing the external task interference and future task changes, simulates the network link target and redundancy requirement based on the model, obtains the initial network topology structure and executes the network link processing, continuously collects the task status data and synchronously corrects the redundancy requirement according to the data to optimize the initial topology, obtains the real-time optimized synchronous network topology, and solves the problem that the static routing and topology optimization methods in the prior art are often unable to cope with the frequently changing network requirements.

[0046] Preferably, the steps of acquiring network structure information and network performance information of a target area network, and performing digital modeling processing on the target area network according to the network structure information and the network performance information to obtain a network simulation model of the target area network include: S11: Acquire network structure information and network performance information of the target area network; wherein the network structure information includes each transmission node of the target area network and the transmission channels between each transmission node, and the network performance information includes node performance characteristics of each transmission node and channel performance characteristics of each transmission channel; S12: performing information analysis of several preset observation and analysis scales on the network structure information and the network performance information to obtain the network description feature distribution of the target area network at each preset observation and analysis scale; S13: performing multi-level digital modeling processing on the target area network according to the distribution of network description characteristics of the target area network at different preset observation and analysis scales, so as to obtain multiple levels of network simulation levels of the target area network; S14: The network simulation layers of multiple levels of the target area network together constitute the network simulation model.

[0047] Specifically, obtain the basic information of all transmission nodes in the target area network, including the name, type, function, etc. of each node. Transmission nodes generally include switches, routers, terminal devices, servers, etc. Obtain the connection mode and link type between each transmission node in the target area network, such as optical fiber link, cable link, wireless link, etc., and obtain the topological connection information of each link, describing which nodes are directly connected and which nodes have intermediate devices such as routers or switches.

[0048] More specifically, the performance indicators of each transmission node are obtained, such as processing power, bandwidth, delay, load capacity, number of ports, etc. Node performance directly affects the processing capacity of network traffic and the response speed of tasks, and the performance characteristics of each transmission channel are obtained, such as bandwidth, delay, packet loss rate, signal-to-noise ratio, etc. These channel characteristics will directly affect data transmission efficiency and task completion time.

[0049] More specifically, different analysis scales are set to analyze the network in detail from different dimensions. Common analysis scales may include: Whole network analysis scale: analyzing the macro characteristics of the entire target area network, understanding the network performance bottlenecks and topological characteristics in a large area, etc. Local area analysis scale: analyzing the performance of a specific area or subnet, such as the performance of a specific node or a specific link, Dynamic time scale: analyzing the performance changes of the network in different time periods, perhaps to identify peak traffic periods, load changes, etc.

[0050] More specifically, according to the set observation and analysis scale, the network structure information and performance information are analyzed, and the distribution of network description features at each scale is extracted. For example, at the scale of the entire network, the total bandwidth, total number of nodes, link fault tolerance, etc. of the network are calculated; at the scale of a local area, the bandwidth, delay, packet loss rate and other performance characteristics of a subnet are analyzed; at the dynamic time scale, the changing trend of network performance over time is analyzed, and the peak load time and bottleneck links of the network are identified.

[0051] More specifically, based on the distribution of network description characteristics of the target area network at different preset observation and analysis scales, multi-level modeling processing is performed. The data at each scale will be used to build network models at different levels. Common modeling levels include: Physical layer modeling: Modeling the physical topology, device performance, and link characteristics of the network, focusing on the basic properties of nodes and links, such as bandwidth, latency, capacity, etc. Logical layer modeling: Based on the physical layer modeling, further modeling is performed on the network traffic, routing strategy, load balancing, etc. The logical layer model focuses more on how data flows through the network and how to allocate network resources. Protocol layer modeling: Consider the impact of network protocols on network performance, such as routing protocols, congestion control mechanisms, QoS strategies, etc. Models at this level will involve the network protocol stack and its optimization of traffic management.

[0052] More specifically, digital modeling techniques are used to model each layer, such as discrete event simulation: used to simulate the occurrence of events in the network, suitable for analyzing dynamic behaviors such as traffic distribution and protocol interaction in the network, graph theory model: the topology of the network is represented by the nodes and edges of the graph, suitable for describing the relationship between nodes and links in the network, and queue theory model: suitable for analyzing queuing behavior and resource competition problems in the network, especially in congestion and high load conditions. Network simulation models at different levels interact with each other to transfer data and status. For example, the link performance of the physical layer will affect the data transmission of the logical layer, and the flow control of the logical layer will affect the routing decision of the protocol layer.

[0053] More specifically, network models at all levels are combined to form a complete network simulation model. This simulation model includes not only detailed modeling of the physical and logical layers, but also strategies and performance optimization of the protocol layer. The modeling results at all levels are integrated to ensure the correctness and consistency of the model. Simulation verification is performed to check whether the network simulation model can accurately reflect the performance and behavior of the target area network. According to the actual test results, the parameters in the model are adjusted and optimized to ensure the accuracy and reliability of the model.

[0054] It is understandable that by obtaining detailed network structure information and performance information and conducting multi-scale analysis, the network simulation model can accurately describe all aspects of the target area network, especially at different preset observation scales, it can comprehensively evaluate various characteristics of the network, such as topology, performance bottlenecks, traffic patterns, etc. Multi-level modeling provides different hierarchical perspectives for network optimization, and can be optimized at different levels. For example, the bandwidth allocation of links can be optimized at the physical layer, the routing strategy can be optimized at the logical layer, and congestion control can be optimized at the protocol layer. The optimizations at each level can work together to improve the overall network performance. The multi-level modeling method makes the network model more flexible and adaptable, and can adapt to different types of network changes. Whether it is a change in network topology or a change in traffic demand, the network model can respond quickly and make corresponding adjustments. By modeling and evaluating network performance at different scales, bottlenecks and potential problems in the network can be accurately identified. For example, the model can identify which nodes or links are performance bottlenecks in the network and which routing strategies may cause increased latency, thereby providing data support for network optimization.

[0055] More specifically, based on the simulation results of digital modeling, network operators can configure network resources more scientifically and accurately. For example, they can predict network loads based on network models, optimize traffic distribution, avoid resource waste, and ensure network stability. Digital modeling provides a virtual environment in which various tests and simulations can be performed, such as traffic surges, node failures, link interruptions, etc. Network administrators can test network performance under different configurations without affecting the actual network and make optimization decisions. Network simulation models can be used to support decision-making for network planning and future expansion. By predicting future network loads and development trends, network planners can formulate expansion strategies in advance to avoid insufficient resources or performance bottlenecks in network expansion.

[0056] Preferably, the steps of obtaining each network link task to be executed by the target area network at the current moment, and setting the network link target for the network simulation model according to each network link task include: S21: performing a real-time task query on the terminal nodes participating in the target area network to receive the network connection requirements of each terminal node at the current moment; S22: Based on the node setting information of each terminal node in the target area network, perform demand execution analysis on the network link demand of each terminal node at the current moment to obtain each network link task to be executed in the target area network at the current moment; S23: Perform multi-dimensional decomposition and priority weight allocation of the execution requirements of the task objectives of each of the network link tasks to obtain the task link objectives of each of the network link tasks, and perform weighted aggregation of the task link objectives to obtain the network link objectives of the target area network.

[0057] Specifically, task inquiries are performed on all terminal nodes in the target area network in real time. Each terminal node (such as user equipment, server or network equipment) will issue a network connection requirement at a certain moment, including bandwidth requirements, delay requirements, specific application protocol support, etc., current network traffic requirements, network quality requirements (such as bandwidth, delay, packet loss rate, etc.), and the timeliness of network tasks (i.e., the time limit requirements for task execution). Through network monitoring tools or the transmission protocols and interfaces of terminal nodes, the network connection requirements from each terminal node at the current moment are received. This information can be collected in real time through API, SNMP protocol, network management system and other means.

[0058] More specifically, according to the node setting information of the target area network (such as bandwidth capacity, load, connection status, etc.), the network link requirements of each terminal node are analyzed. In this step, the current status of the network node (for example, the bandwidth or load of some nodes may have reached the limit) will affect whether the task can be successfully executed. The network link requirements of each terminal node are analyzed in detail. The specific analysis content includes: whether there is a network resource bottleneck, such as whether the bandwidth of the link is sufficient, whether the processing capacity of the node can meet the demand, whether the link between the nodes is stable enough to ensure the smooth progress of the link task, whether it is necessary to give priority to certain urgent tasks or high-priority link requirements. According to the results of the demand analysis, the tasks to be executed in the current network are classified and the corresponding network link tasks are generated. Each task will include the target of the network connection, bandwidth requirements, delay requirements, time limit requirements, etc.

[0059] More specifically, for each network link task, multi-dimensional decomposition is performed to identify multiple dimensional requirements in the task. Common decomposition dimensions include: bandwidth requirements: the bandwidth demand of each task, timeliness requirements: the time limit or priority of the task, especially high-priority tasks need to be executed in advance, delay requirements: the maximum delay range required for the network connection, reliability requirements: for some tasks, low packet loss rate or high reliability may be required.

[0060] More specifically, each task is assigned a priority. Common criteria for priority assignment include: task urgency: high-real-time applications such as real-time video calls, VoIP, and online games have higher priority; task importance: certain key businesses (such as payment systems, medical equipment data transmission, etc.) have higher priority.

[0061] More specifically, if a task has a high bandwidth requirement and network bandwidth resources are tight, it may be necessary to assign a lower priority to the task or assign it to a link with sufficient bandwidth. The priority of each task can be quantified through a weight model. For example, a weighted average method or a machine learning model can be used to automatically adjust the weight of the task to ensure that the most important and urgent tasks are given priority.

[0062] More specifically, according to the decomposition and priority weight allocation in the previous steps, the network link target of each task is defined. This target can be a comprehensive value, which represents the comprehensive requirements of bandwidth, delay, reliability and other characteristics required by the task. The link targets of all tasks are weighted and summarized according to the priority of the task. At this time, the priority weight of each task will affect the final summary result. The summarized target will be the network resource configuration and optimization target required by the target area network at the current moment. For example, the weighted bandwidth requirements, optimal delay time, network load balancing, etc. of all tasks can be summarized into a total network link target. This target reflects the optimal link requirements of the network at the current moment. The summarized task link targets can be directly used as input for the network simulation model to perform optimization calculations, or as the basis for subsequent network resource allocation.

[0063] It can be understood that by real-time task inquiries to terminal nodes, accurate task requirement information can be obtained to ensure that the allocation of network resources meets the actual needs of each task. This step helps to dynamically identify and adjust link tasks in the network and adapt to the network performance requirements of different applications. Task decomposition and priority allocation enable the network to prioritize the most urgent and important network tasks under high load or resource constraints. This intelligent allocation mechanism helps to avoid task delays and improve the overall performance and response speed of the network. By weighted aggregation of task goals, network resource configuration (such as bandwidth, delay, reliability, etc.) can be aggregated into an optimization goal, so that the network can provide the best link service at the current moment. This optimization can help network managers allocate resources reasonably and ensure efficient network operation. The aggregated network link goals can be used as input data for real-time optimization of the network simulation model at the current moment. This provides network operators with a flexible response plan that can quickly respond to network traffic fluctuations, node failures or other emergencies.

[0064] Preferably, the steps of performing a predictive analysis of external task interference and future task changes on the network simulation model, and setting performance redundancy requirements for the network simulation model according to the analysis results include: S31: Acquire network tag information of the target area network and time tag information of the current moment, and combine the network tag information and event tag information to obtain prediction object information; S32: performing a predictive analysis of external task interference on the prediction object information based on a pre-trained neural network model to obtain external task interference prediction features of the network simulation model; S33: performing task feature identification on each of the network link tasks to obtain a task identification tag of each of the network link tasks, and performing search and matching of historical data on the network link tasks according to the task identification tag based on a historical database to obtain historical reference features of the network link tasks; wherein the historical reference features include duration features, change tendency features, and new task growth trend features; S34: performing a predictive analysis of future task changes on the network link target according to the historical reference features to obtain a future task change prediction feature of the network link target in a future time period; S35: Performing a network performance demand analysis on the network simulation model based on preset response standards according to the external task interference prediction characteristics and the future task change prediction characteristics to obtain the performance redundancy requirements of the network simulation model, and configuring model parameters of the network simulation model according to the performance redundancy requirements.

[0065] Specifically, network tag information and time tag information are obtained. The network tag information includes the status, load, bandwidth, delay and other related information of each node and link in the target area network. This information provides detailed data of the current network status. The time tag information refers to the timestamp or time interval of the current time. Time tag information is very important for capturing the trend of network status changing over time and periodic fluctuations. The network tag information and time tag information are combined to form the prediction object information. This combination can help accurately understand the network status and potential interference sources at a specific time point.

[0066] More specifically, the neural network model is trained through historical data and the execution of network tasks to enable it to identify and predict the interference characteristics of external tasks (such as sudden increase in traffic, external attacks, unexpected failures, etc.), and the combined prediction object information is input. Through the neural network model analysis, the interference characteristics of external factors on the network in the current and future time periods are predicted. The prediction results will include: external task interference type: such as traffic surge, unexpected events, external service impact, etc., interference intensity: for example, changes in indicators such as bandwidth occupancy, delay increase, packet loss rate, etc., interference duration: the time range in which the interference may affect the network.

[0067] More specifically, the task characteristics of each network link task are identified, which usually include: bandwidth demand: the bandwidth demand of the task, delay demand: the delay requirement of the task, priority: the execution priority of the task, which may affect the scheduling of network resources; based on the task identification tag, the system finds similar historical task characteristics by querying the task execution data in the historical database. The goal of historical data matching is to identify the performance of tasks in similar situations: duration characteristics: the duration of task execution, which helps to infer the stability and continuous demand of task execution, change tendency characteristics: the change trend of task performance (such as bandwidth, delay), through the analysis of historical data, predict the possible fluctuation or growth of the task, new task growth trend characteristics: identify the growth pattern of new tasks, which may be sudden, gradual, or seasonal.

[0068] More specifically, based on historical reference features (duration, tendency to change, and growth trend of new tasks), predictions are made on future task changes. The goal of the prediction is to estimate changes in tasks in future time periods: task bandwidth growth: whether the bandwidth demand for tasks is expected to increase, whether there is a possibility of sudden tasks, latency fluctuations: whether the latency demand for tasks will change significantly, and task distribution changes: whether the distribution of tasks among different network nodes has changed.

[0069] More specifically, the prediction features of the final output include: the growth of task demand: predicting whether the demand for task bandwidth, latency, etc. will increase or decrease in a certain period of time in the future; changes in task characteristics: the changing trend of task characteristics, such as whether new tasks will be added, and whether the bandwidth and latency of existing tasks will fluctuate.

[0070] More specifically, the performance requirements analysis of the network simulation model is carried out by combining the predicted characteristics of external task interference with the predicted characteristics of future task changes. The goal of this analysis is to ensure that the network can maintain its efficient performance under possible interference and task changes. The impact of external interference, such as the bandwidth competition caused by the increase of external tasks, or the traffic peak caused by external attacks, and the adaptability to future task changes: whether the network has sufficient redundant resources to cope with the increase or decrease of task requirements.

[0071] More specifically, based on the above analysis results, the performance redundancy requirements of the network model are set. The redundancy requirements refer to the additional resources or redundant capabilities required by the network under normal operation and external interference, task changes, such as: redundant bandwidth requirements: increase a certain proportion of bandwidth to ensure sufficient network resources during peak hours or bursts of traffic, redundant delay: set a certain delay tolerance to ensure that the delay in task execution does not exceed the preset standard, link redundancy: when certain links are interfered with, they can automatically switch to backup links to ensure the stability of network connections, and strengthen fault-tolerant mechanisms: such as the fault-tolerant capabilities of network nodes, load balancing mechanisms, etc.

[0072] More specifically, according to the performance redundancy requirements, the parameter configuration of the network simulation model is adjusted, including: bandwidth adjustment: adjusting the bandwidth resource allocation in the model according to the redundancy requirements, delay limit: setting the delay limit in the network simulation model to ensure that the task is completed within the specified time, resource reservation strategy: reserving a certain amount of resources to deal with external interference and task fluctuations to avoid excessive tension in network resources.

[0073] It is understandable that the neural network model can identify and predict the impact of external factors on network performance, provide network administrators with timely interference warnings, and effectively reduce the impact of sudden network interference on service quality. Through historical data matching and task feature identification, it can predict future changes in tasks and help the network optimize task scheduling, load balancing and resource allocation. By predicting task changes and external interference, the network can automatically adjust resource configuration and redundancy requirements, and enhance the network's ability to respond to changes, especially in high-load or complex task environments. Predictive analysis provides a basis for redundant resource configuration, reduces resource waste, and ensures timely response when network resources are tight or interference occurs, thereby improving network stability and reliability.

[0074] More specifically, based on real-time predictions, the network simulation model can be dynamically adjusted and automatically optimize parameter configurations to improve the network's emergency response speed and resource utilization efficiency. By combining external interference predictions and task change predictions, the network can achieve more intelligent self-adjustment, thereby reducing the burden of manual intervention and improving the level of automation and intelligence in network management. Accurate predictions and redundant requirement settings can effectively reduce network risks caused by external interference or task fluctuations, and reduce the probability of service interruptions and performance degradation.

[0075] Preferably, the steps of performing network link simulation on the network link target and the performance redundancy requirement based on the network simulation model to obtain an initial link topology structure of the target area network, and performing network link processing on the target area network according to the initial link topology structure include: S41: performing a network link simulation adapted to the network link target on the network simulation model having multiple levels of network simulation levels through a self-organizing network topology algorithm, and synchronously supervising the simulation target during the network link simulation process according to the performance redundancy requirement, so as to obtain several preliminary link topology structures of the target area network; S42: Analyzing the executability of the distribution optimization algorithm for the various preliminary link topologies to obtain the execution feasibility of the distribution optimization algorithm for the various preliminary link topologies, and determining the initial link topology among the various preliminary link topologies by comparative analysis of the execution feasibility of the distribution optimization algorithm for the various preliminary link topologies; S43: Analyzing the decision nodes of the optimization algorithm on the initial link topology structure according to the distribution optimization algorithm to obtain edge decision nodes and core execution nodes in the initial link topology structure; S44: Configure node working parameters of each transmission node of the target area network according to the edge decision nodes and the core execution nodes in the initial link topology structure, deploy the node network of edge decision nodes and core execution nodes in the target area network, and construct transmission channels between the edge decision nodes and the core execution nodes, so that the target area network is in a network link state corresponding to the initial link topology structure.

[0076] Specifically, a self-organizing network topology algorithm (such as an adaptive topology construction algorithm based on local rules) is used to simulate the target area network. The algorithm can adaptively adjust the network topology according to the distribution of network nodes, task requirements, and external interference information to adapt to different network link targets. The network simulation model is divided into multiple levels, and the network simulation level can include link layer, node layer, and transmission layer. The algorithm simulates different network topologies by self-organizing adjustments at each level. During the simulation process, performance redundancy requires synchronous supervision of the simulated targets. That is, during the simulation process, ensure that the adjustment of the network topology meets the redundancy requirements to prevent single point failures or performance degradation in the network. Based on the results of the self-organizing algorithm simulation, multiple possible network topologies are generated, and each prepared topology is one of the best network structures that may appear under different interference and task demand changes.

[0077] More specifically, a distributed optimization algorithm (such as genetic algorithm, particle swarm optimization algorithm, etc.) is used to evaluate each preliminary topology structure. The distributed optimization algorithm is used to optimize the resource allocation of each node and the load balancing of the link. By analyzing the execution of the distributed optimization algorithm in each preliminary topology structure, the optimization effect and the feasibility of actual execution of each topology structure are judged. The specific analysis content includes: the rationality of link load distribution, the effectiveness of node resource allocation, whether bottlenecks and single point failures can be effectively avoided, and other problems. The optimization results of different topologies are compared to find the most suitable topology structure under the current redundancy requirements.

[0078] More specifically, based on the feasibility analysis of various prepared topologies, comparisons are made and the optimal topology is determined. The optimal topology refers to a structure that can maximize network performance (such as bandwidth, latency, load balancing, etc.) and has high feasibility while meeting redundancy requirements. By comparing the advantages and disadvantages of different prepared topologies, an initial network topology is ultimately determined, which can meet the optimal performance of the target area network under specific mission requirements and redundancy requirements.

[0079] More specifically, the determined initial topology is further analyzed to identify key optimization algorithm decision nodes, which are usually core nodes or decision nodes that affect network performance, such as edge nodes, core nodes, etc. Edge decision nodes and core execution nodes: Edge decision nodes: nodes located at the edge of the network, responsible for processing user requests or data transmission. The decisions of these nodes have an important impact on the performance of the network. Core execution nodes: nodes in the core part of the network, responsible for the scheduling and execution of core tasks, usually require stronger computing and storage capabilities. The purpose of the optimization algorithm decision node analysis is to determine which nodes in the network undertake key tasks in the topology structure. The optimization of these nodes directly affects the stability, efficiency and redundancy of the network.

[0080] More specifically, according to the analysis results of the optimization algorithm decision node, appropriate working parameters are configured for each network node. These parameters include bandwidth allocation, delay optimization, node load, etc., to ensure that each node in the network can work efficiently and stably, and the deployment of edge decision nodes and core execution nodes: edge nodes are responsible for accessing users or terminals, and may require lower latency and higher access bandwidth. Core nodes are responsible for the core processing and scheduling of the network, and usually require stronger processing power and fault tolerance. According to the role and task requirements of the node, its corresponding working parameters are configured, such as processing power, bandwidth, delay limit, etc.

[0081] More specifically, reliable transmission channels are established between edge decision nodes and core execution nodes. These channels need to support high bandwidth and low latency requirements to ensure efficient data transmission in the network. In order to prevent the occurrence of single point failures, the transmission channel design between edge nodes and core nodes in the network needs to have redundancy, that is, multiple channels work in parallel to ensure that when a channel fails, the network can still maintain stable operation, and ensure that all network nodes and links are in the expected network connection state based on the initial link topology. Through self-organizing topology and redundant design, the network can automatically adapt to external interference and task changes and maintain stable operation.

[0082] It can be understood that through the self-organizing network topology algorithm, the network can automatically adapt to the changes in the target area and task requirements, generate multiple preliminary topology structures to cope with network changes in different scenarios, use the distributed optimization algorithm to evaluate the preliminary topology structure, and through feasibility analysis, ensure the selection of the optimal topology structure, which can not only improve bandwidth utilization, but also reduce latency, avoid congestion and bottlenecks, and ensure that the network has sufficient fault tolerance by optimizing the redundant design to avoid network interruption or performance degradation due to single point failure. The redundant design of edge nodes and core nodes ensures the stability and reliability of the network.

[0083] More specifically, the working parameters of network nodes are accurately configured according to the roles and task requirements of the nodes to improve the resource utilization and processing capacity of the network. At the same time, the response speed and execution efficiency of the network are guaranteed by the reasonable deployment of edge nodes and core nodes, and an efficient and low-latency transmission channel is established between edge nodes and core nodes to ensure the efficiency of data transmission, especially in high-traffic or task-intensive situations. The stability and rapid response of the network can be guaranteed. This process can automatically adjust the network topology and resource allocation through self-organizing algorithms, optimized decisions and node configuration, reduce manual intervention, and improve the intelligence and automation level of network management.

[0084] Preferably, the steps of performing network link simulation adapted to the network link target on the network simulation model having multiple levels of network simulation levels by using a self-organizing network topology algorithm, and synchronously supervising the simulation target on the process of network link simulation according to the performance redundancy requirement to obtain several preliminary link topology structures of the target area network include: S411: progressively disassemble the multiple layers of network simulation hierarchies of the network simulation model to obtain a hierarchical simulation sequence consisting of a plurality of hierarchical simulation models arranged in sequence from a high-level generalization to a low-level specificity; S412: performing a progressive simulation operation adapted to the network link target on each of the hierarchical simulation models in the hierarchical simulation sequence repeatedly for multiple times by using any one of a genetic algorithm, an ant colony algorithm and a particle swarm optimization algorithm, wherein the progressive simulation operation process includes first performing a simulation operation on a higher-level generalized hierarchical simulation model, and then performing a simulation operation on a lower-level specific hierarchical simulation model based on the result of the simulation operation, so as to obtain a plurality of preliminary link topological structures that meet the network link target; S413: Synchronously monitoring the optimization direction of the simulation operation during the progressive simulation algorithm according to the performance redundancy requirement, so that the various preparation link topologies meet the performance redundancy requirement.

[0085] Specifically, the target network simulation model is decomposed into multiple levels of simulation models, which are progressive from high level to low level. Each layer represents a different level of network abstraction. The upper-level model describes the network's macro structure, global behavior and rough performance, while the lower-level model describes the specific details of the network, the configuration and optimization of nodes and links. Through this progressive decomposition, it is ensured that the preliminary results obtained in the high-level model can guide the more detailed simulation calculations of the lower level, so that the optimization decisions of each layer are globally consistent. These simulation models at different levels are arranged in a "top-down" order to form a hierarchical simulation sequence. In this way, the optimization process gradually deepens from macro to micro, ensuring that the network topology can obtain the best configuration at different levels.

[0086] More specifically, multiple iterative operations are performed through any one of the genetic algorithms, ant colony algorithms or particle swarm optimization algorithms. These algorithms are suitable for global optimization, can handle complex network topologies and performance requirements, and find the global optimal solution or approximate optimal solution. First, the upper model in the hierarchical simulation sequence is simulated. This stage is usually a more general simulation, dealing with a larger network range and high-level network behaviors, such as the layout of network nodes and the basic connection of links. According to the operation results of the upper-level model, the lower-level models are simulated and optimized step by step. This process gradually adjusts the specific parameters of each node (such as bandwidth, delay, etc.) and link configuration so that the network can be optimized at different levels. The simulation operation results of each level will affect the simulation input of the next level. This process is iterative until a preliminary link topology structure that meets the network link target is obtained. Through multiple rounds of simulation, the network topology will gradually approach the optimal solution.

[0087] More specifically, during the process of progressive simulation, the optimization direction of the simulation operation is monitored in real time to ensure that the optimization at each level takes into account the performance redundancy requirements. The redundancy requirements refer to how to configure additional nodes, links, and paths in the network topology to ensure that the network can continue to operate in the event of a failure and avoid single point failures or overloads. In each round of optimization, the redundancy is monitored to ensure that even if a part of the links or nodes fails, other links or nodes can take over their functions, thereby ensuring the stability and high availability of the network. Synchronous supervision can avoid the risk of network paralysis due to insufficient redundancy.

[0088] More specifically, through progressive simulation and redundant supervision, multiple different preliminary link topologies can be generated. These topologies are adapted to different network goals and redundancy requirements. Each topology will be adjusted according to different network requirements (such as bandwidth, latency, load balancing, etc.) during the optimization process to ensure that the network can perform optimally under different circumstances. These preliminary topologies will provide a variety of options to facilitate the subsequent selection of the topology that best suits specific task requirements.

[0089] It is understandable that through the self-organizing network topology algorithm, the network simulation process can dynamically adjust the topology structure according to different goals and task requirements to ensure that the network can adapt to various situations such as external changes and load fluctuations. The use of a progressive simulation algorithm helps to refine the optimization of the network topology layer by layer, so that the simulation of each layer can provide reasonable constraints and optimization goals for the next layer. This progressive method can avoid ignoring micro details during global optimization, thereby improving the overall efficiency and stability of the network. During the optimization process, the redundant design is synchronously supervised to ensure that each prepared topology structure can meet the fault tolerance requirements of the network and avoid the occurrence of single point failures. The redundant design improves the reliability of the network and ensures high availability.

[0090] More specifically, by introducing global optimization algorithms such as genetic algorithms, ant colony algorithms, or particle swarm optimization algorithms, it is possible to ensure that the network topology achieves optimal configuration globally, while local optimization is adjusted at each level of detail to ensure the network's excellent performance in multiple indicators. By generating multiple preparatory topologies, the most suitable topology can be flexibly selected according to different application scenarios and requirements. This flexibility provides more options for subsequent network design and optimization, and can adapt to different types of business requirements and network conditions. By adding real-time supervision of performance redundancy during the simulation process, it ensures that the network can still maintain stable performance and high availability in the face of sudden failures or high loads. This optimization ensures the long-term operational stability of the network.

[0091] Preferably, the target area network continuously collects the real-time status of the task to obtain the real-time characteristics of the task, and synchronously corrects the performance redundancy requirements of the network simulation model according to the real-time characteristics of the task, so as to drive the network simulation model to perform real-time optimization on the initial link topology structure to obtain the synchronous link topology structure of the target area network. The step of performing network link processing on the target area network includes: S51: continuously collecting real-time information of the network link task of the target area network through the pre-deployed network monitoring sensor unit to obtain the real-time task characteristics of the target area network; S52: performing a difference analysis on the network link targets of the network simulation model according to the actual task characteristics, and evaluating the expected compliance degree of the difference analysis results with the performance redundancy requirements to obtain a redundancy value parameter of the performance redundancy requirements; S53: judging whether to retain, expand or reduce the performance redundancy requirement according to the redundancy value parameter to obtain a correction action of the performance redundancy requirement, and executing the performance redundancy requirement according to the correction action to obtain a performance redundancy requirement that meets the actual characteristics of the task; wherein the correction action includes redundancy retention, redundancy expansion or redundancy reduction; S54: Based on the current network link target and the revised performance redundancy requirement, the initial link topology structure of the network simulation model is optimized in real time to obtain the synchronous link topology structure of the target area network, and the network link processing of the target area network is performed according to the synchronous link topology structure.

[0092] Specifically, network monitoring sensor units (such as sensors, monitoring equipment, etc.) are pre-deployed in the target area network. These devices are responsible for continuously monitoring the real-time status of the network and the operation of tasks. These sensor units will continuously collect the real-time characteristics of network tasks, including network operation indicators such as traffic, bandwidth, delay, packet loss rate, node load, etc. The collected real-time information will be used to analyze the execution status and characteristics of the current network tasks to obtain the "real-time characteristics" of the network tasks. These characteristics reflect the working status of the current network when carrying specific tasks, such as traffic fluctuations, resource consumption, link stability, etc.

[0093] More specifically, based on the collected actual task characteristics, we first compare and analyze these characteristics with the network link targets preset in the network simulation model. The purpose of the difference analysis is to find out the differences between the actual performance of the current network and the expected targets of the simulation model, such as inconsistent bandwidth requirements, excessive node load or insufficient redundancy.

[0094] More specifically, based on the results of the difference analysis, evaluate whether the current network meets the expected performance redundancy requirements. If the network's redundancy is insufficient, it may lead to unstable task execution or network bottlenecks; if the redundancy is too high, resources may be wasted. Through difference analysis, the "redundancy value parameters" of the redundancy requirements can be quantified. These parameters reflect whether the current network needs to increase or decrease redundancy.

[0095] More specifically, based on the redundancy value parameters, it is determined whether the performance redundancy requirements of the network need to be revised. The correction actions can be: Redundancy retention: If the current redundancy meets the task requirements, the existing redundancy configuration is retained to ensure network stability; Redundancy expansion: If the task characteristics indicate insufficient redundancy (for example, the network load is too high or the redundant links are insufficient), redundancy expansion is performed to add more backup links or resources; Redundancy reduction: If the network resources are overly redundant, resulting in resource waste (for example, the redundancy of some links or nodes is too high), redundancy is reduced to optimize resource allocation.

[0096] More specifically, based on the above judgment, corresponding corrective actions are implemented. If it is decided to expand or reduce redundancy, the system will adjust the network's redundancy configuration and modify the allocation of related nodes, links or resource pools to ensure that the network redundancy meets the expected performance standards.

[0097] More specifically, combined with the revised performance redundancy requirements, the initial link topology is optimized in real time through a network simulation model. The goal of the optimization is to adapt the network topology to the current task requirements and ensure that the task can be executed stably and efficiently. This optimization may involve link reconfiguration, node reconnection, or the addition of new redundant links to ensure that the network has sufficient fault tolerance and load balancing capabilities to avoid bottlenecks and failures.

[0098] More specifically, a synchronous link topology structure is formed through the optimized network topology. This structure is the result of dynamic adjustment based on real-time task requirements and performance redundancy requirements. It can provide optimal resource allocation and link planning. The synchronous link topology structure ensures that the network has optimal connectivity and redundancy capabilities in the current state, thereby ensuring efficient execution of tasks.

[0099] More specifically, based on the optimized synchronous link topology, the actual link processing of the target area network is carried out, which includes the connection between network nodes, resource allocation, link setting, etc., so that the network can operate according to the latest topology and redundancy requirements and meet the performance and stability requirements of the current task.

[0100] It is understandable that through continuous task real-time information collection and real-time analysis, it is possible to ensure that the optimization of network topology is adapted to task requirements and adjust network configuration in real time. This dynamic adaptability ensures that the network can cope with various changes and challenges that may arise during task execution. The synchronous correction of redundancy requirements based on the characteristics of the task real-time effectively improves the fault tolerance and stability of the network. Through redundancy expansion or reduction, the network can adapt to the load requirements of different tasks while avoiding the waste of resources caused by excessive redundancy.

[0101] More specifically, the redundancy configuration of the network is precisely adjusted through difference analysis and redundant value parameter evaluation. The corrective actions of redundancy retention, expansion or reduction can ensure that the network achieves the optimal configuration of resources while meeting the mission requirements, avoiding the problem of insufficient or excessive redundancy. Through real-time optimization of the initial link topology, the network topology is always in the best state, maximizing the execution efficiency of the mission and reducing network problems such as delay and packet loss, which enables the network to better support high-load and high-demand tasks.

[0102] Reference Figure 2 As shown, in a second aspect, the present invention provides an intelligent routing optimization system in a high-speed optical network, which is used to implement an intelligent routing optimization method in a high-speed optical network as described in any one of the first aspects, including: A network simulation module, used to obtain network structure information and network performance information of a target area network, and perform digital modeling processing on the target area network according to the network structure information and the network performance information to obtain a network simulation model of the target area network; A target setting module, used for obtaining each network link task to be executed by the target area network at the current moment, and setting the network link target for the network simulation model according to each network link task; A redundancy setting module, used to perform a predictive analysis of external task interference and future task changes on the network simulation model, and to set performance redundancy requirements for the network simulation model according to the analysis results; A link simulation module, configured to perform network link simulation on the network link target and the performance redundancy requirement based on the network simulation model to obtain an initial link topology structure of the target area network, and perform network link processing on the target area network according to the initial link topology structure; A real-time optimization module is used to continuously collect the real-time status of the task status of the target area network to obtain the real-time characteristics of the task, and to synchronously correct the performance redundancy requirements of the network simulation model according to the real-time characteristics of the task, so as to drive the network simulation model to perform real-time optimization on the initial link topology structure, so as to obtain the synchronous link topology structure of the target area network and perform network link processing on the target area network.

[0103] In this embodiment, for the specific implementation of each module in the above system embodiment, please refer to the above method embodiment, which will not be repeated here.

[0104] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. An intelligent routing optimization method in a high-speed optical network, characterized in that: include: Acquiring network structure information and network performance information of a target area network, and performing digital modeling processing on the target area network according to the network structure information and the network performance information to obtain a network simulation model of the target area network; Acquire each network link task to be executed by the target area network at the current moment, and set the network link target for the network simulation model according to each network link task; Performing a predictive analysis of external task interference and future task changes on the network simulation model, and setting performance redundancy requirements for the network simulation model based on the analysis results; Performing network link simulation on the network link target and the performance redundancy requirement based on the network simulation model to obtain an initial link topology structure of the target area network, and performing network link processing on the target area network according to the initial link topology structure; The target area network continuously collects the real-time status of the mission to obtain the real-time characteristics of the mission, and synchronously corrects the performance redundancy requirements of the network simulation model according to the real-time characteristics of the mission, so as to drive the network simulation model to perform real-time optimization on the initial link topology structure, so as to obtain the synchronous link topology structure of the target area network and perform network link processing on the target area network.

2. The intelligent routing optimization method in a high-speed optical network according to claim 1, characterized in that: The steps of obtaining network structure information and network performance information of a target area network, and performing digital modeling processing on the target area network according to the network structure information and the network performance information to obtain a network simulation model of the target area network include: Acquire network structure information and network performance information of the target area network; wherein the network structure information includes each transmission node of the target area network and the transmission channels between each transmission node, and the network performance information includes the node performance characteristics of each transmission node and the channel performance characteristics of each transmission channel; Performing information analysis of several preset observation and analysis scales on the network structure information and the network performance information to obtain the network description feature distribution of the target area network at each preset observation and analysis scale; According to the distribution of network description characteristics of the target area network at different preset observation and analysis scales, the target area network is subjected to multi-level digital modeling processing to obtain multiple levels of network simulation levels of the target area network; The target area network's multiple layers of network simulation levels together constitute the network simulation model.

3. The intelligent routing optimization method in a high-speed optical network according to claim 1, characterized in that: The steps of obtaining each network link task to be executed by the target area network at the current moment and setting the network link target for the network simulation model according to each network link task include: Performing real-time task inquiries on terminal nodes participating in the target area network to receive network connection requirements of each terminal node at the current moment; Based on the node setting information of each terminal node in the target area network, performing demand execution analysis on the network link demand of each terminal node at the current moment, so as to obtain each network link task to be executed in the target area network at the current moment; The execution requirements of the task objectives of each of the network link tasks are multi-dimensionally decomposed and priority weights are allocated to obtain the task link objectives of each of the network link tasks, and the task link objectives are weighted and summarized to obtain the network link objectives of the target area network.

4. The intelligent routing optimization method in a high-speed optical network according to claim 1, characterized in that: The steps of performing a predictive analysis of external task interference and future task changes on the network simulation model, and setting performance redundancy requirements for the network simulation model according to the analysis results include: Acquire network tag information of the target area network and time tag information of the current moment, combine the network tag information and event tag information to obtain prediction object information; Performing a predictive analysis of external task interference on the prediction object information based on a pre-trained neural network model to obtain external task interference prediction features of the network simulation model; Performing task feature identification on each of the network link tasks to obtain a task identification tag for each of the network link tasks, and searching and matching historical data of the network link tasks based on the task identification tags based on a historical database to obtain historical reference features of the network link tasks; wherein the historical reference features include duration features, change tendency features, and new task growth trend features; Performing a predictive analysis of future task changes on the network link target according to the historical reference features to obtain a future task change prediction feature of the network link target in a future time period; According to the external task interference prediction characteristics and the future task change prediction characteristics, the network performance demand analysis of the network simulation model based on preset response standards is performed to obtain the performance redundancy requirements of the network simulation model, and the model parameters of the network simulation model are configured according to the performance redundancy requirements.

5. The intelligent routing optimization method in a high-speed optical network according to claim 2, characterized in that: The steps of performing network link simulation on the network link target and the performance redundancy requirement based on the network simulation model to obtain an initial link topology structure of the target area network, and performing network link processing on the target area network according to the initial link topology structure include: Performing a network link simulation adapted to the network link target on the network simulation model having multiple levels of network simulation levels through a self-organizing network topology algorithm, and synchronously supervising the simulation target during the network link simulation process according to the performance redundancy requirement, so as to obtain several kinds of preliminary link topology structures of the target area network; Analyzing the executability of the distribution optimization algorithm on the various preliminary link topologies to obtain the feasibility of the execution of the distribution optimization algorithm corresponding to the various preliminary link topologies, and determining the initial link topology among the various preliminary link topologies by comparing and analyzing the feasibility of the execution of the distribution optimization algorithm corresponding to the various preliminary link topologies; Analyzing the decision nodes of the optimization algorithm on the initial link topology structure according to the distributed optimization algorithm to obtain edge decision nodes and core execution nodes in the initial link topology structure; According to the edge decision nodes and the core execution nodes in the initial link topology structure, the node working parameters of each transmission node of the target area network are configured to perform node network deployment of edge decision nodes and core execution nodes in the target area network, and to construct transmission channels between the edge decision nodes and the core execution nodes, so that the target area network is in a network link state corresponding to the initial link topology structure.

6. The intelligent routing optimization method in a high-speed optical network according to claim 5, characterized in that: The steps of performing network link simulation adapted to the network link target on the network simulation model having multiple levels of network simulation levels by using a self-organizing network topology algorithm, and synchronously supervising the simulation target on the process of network link simulation according to the performance redundancy requirement to obtain several preliminary link topology structures of the target area network include: The multiple layers of network simulation hierarchies of the network simulation model are progressively decomposed to obtain a hierarchical simulation sequence consisting of a plurality of hierarchical simulation models arranged in sequence from a high-level generalization to a low-level specificity; Repeating multiple times any one of a genetic algorithm, an ant colony algorithm and a particle swarm optimization algorithm to perform progressive simulation operations adapted to the network link target on each of the hierarchical simulation models in the hierarchical simulation sequence, wherein the progressive simulation operation process includes first performing simulation operations on the upper-level generalized hierarchical simulation model, and then performing simulation operations on the lower-level specific hierarchical simulation models based on the results of the simulation operations, so as to obtain several preliminary link topological structures that meet the network link target; According to the performance redundancy requirement, the optimization direction of the simulation operation is synchronously supervised during the progressive simulation algorithm so that the various preparation link topologies meet the performance redundancy requirement.

7. The intelligent routing optimization method in a high-speed optical network according to claim 1, characterized in that: The target area network continuously collects the real-time status of the task to obtain the real-time characteristics of the task, and synchronously corrects the performance redundancy requirements of the network simulation model according to the real-time characteristics of the task, so as to drive the network simulation model to perform real-time optimization on the initial link topology structure to obtain the synchronous link topology structure of the target area network. The step of performing network link processing on the target area network includes: The real-time information of the network link task is continuously collected from the target area network through the pre-deployed network monitoring sensor unit to obtain the real-time task characteristics of the target area network; Performing a difference analysis on the network link targets of the network simulation model according to the actual characteristics of the task, and evaluating the expected compliance degree of the difference analysis results with the performance redundancy requirements to obtain a redundancy value parameter of the performance redundancy requirements; According to the redundancy value parameter, the performance redundancy requirement is judged to retain, expand or reduce redundancy, so as to obtain a correction action of the performance redundancy requirement, and the performance redundancy requirement is executed according to the correction action, so as to obtain a performance redundancy requirement that meets the actual characteristics of the task; wherein the correction action includes redundancy retention, redundancy expansion or redundancy reduction; Based on the current network link target and the revised performance redundancy requirement, the initial link topology of the network simulation model is optimized in real time to obtain the synchronous link topology of the target area network, and network link processing is performed on the target area network according to the synchronous link topology.

8. An intelligent routing optimization system in a high-speed optical network, characterized in that: A method for implementing an intelligent routing optimization method in a high-speed optical network as described in any one of claims 1 to 7, comprising: A network simulation module, used to obtain network structure information and network performance information of a target area network, and perform digital modeling processing on the target area network according to the network structure information and the network performance information to obtain a network simulation model of the target area network; A target setting module, used for obtaining each network link task to be executed by the target area network at the current moment, and setting the network link target for the network simulation model according to each network link task; A redundancy setting module, used to perform a predictive analysis of external task interference and future task changes on the network simulation model, and to set performance redundancy requirements for the network simulation model according to the analysis results; A link simulation module, configured to perform network link simulation on the network link target and the performance redundancy requirement based on the network simulation model to obtain an initial link topology structure of the target area network, and perform network link processing on the target area network according to the initial link topology structure; A real-time optimization module is used to continuously collect the real-time status of the task status of the target area network to obtain the real-time characteristics of the task, and to synchronously correct the performance redundancy requirements of the network simulation model according to the real-time characteristics of the task, so as to drive the network simulation model to perform real-time optimization on the initial link topology structure, so as to obtain the synchronous link topology structure of the target area network and perform network link processing on the target area network.

Images