A multi-network information management system based on artificial intelligence

By using multiple network information management systems based on artificial intelligence in a high-traffic network environment, monitoring and analysis of traffic and operation risks, dynamically selecting switching modes and migration strategies, the data during network link switching in the existing technology is solved. Integrity and business continuity issues achieve efficient and reliable traffic migration and resource conservation.

CN118740699BActive Publication Date: 2024-11-15XIAN KANGNAI NETWORK TECH CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411233396.4
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-09-04
Publication Date
2024-11-15
Estimated Expiration
2044-09-04

AI Technical Summary

Technical Problem

In a high-traffic network environment, the existing technology lacks attention to operational risks during network link switching, which may lead to loss of data packets or incorrect transmission, affecting data integrity and business continuity, and at the same time, there is a problem of migration mode solidification. resulting in waste of resources and time costs.

Method used

Adopt multiple network information management systems based on artificial intelligence, and operate monitoring modules, traffic migration requirements identification modules, health link screening modules, adaptive migration mode determination modules, single-link gradual migration modules and dynamic adjustment of traffic allocation through network links The module conducts synchronous monitoring of traffic and operational risks, identifying excessive links and healthy links, analyzing traffic distribution characteristics and migration support, and dynamically selecting switching modes and migration strategies.

Benefits of technology

It improves the efficiency and reliability of network link handover, minimizes data loss and service interruption during the handover process, ensures the integrity and business continuity of data transmission, and saves resources and time costs.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118740699B_ABST
    Figure CN118740699B_ABST
Patent Text Reader

Abstract

The present invention belongs to the field of communication technology, and specifically discloses a variety of network information management systems based on artificial intelligence. By monitoring traffic and operation risks during the operation of multiple network links, and identifying demand links and healthy links based on the monitoring results, at the same time, by analyzing the traffic distribution characteristics of the demand link and the traffic migration support of the healthy link, an adaptive migration mode is selected, which reflects the targeted and adaptive migration of traffic in network link switching, and can maximize the dual guarantees of switching effect and saving switching resources on the basis of improving switching efficiency. Then, after selecting the adaptive migration mode, the traffic is gradually and dynamically migrated according to the traffic and operation risk monitoring results of the links associated with the switching, which effectively avoids the one-time switching of the link under the operation risk state, greatly improves the integrity and continuity of data transmission during the switching process, and is conducive to the effective and reliable implementation of network link switching.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention belongs to the field of communication technology, and specifically discloses a multiple network information management system based on artificial intelligence. Background Art

[0002] With the popularization of the Internet and the increase in smart devices, users' online activities and data demands have increased significantly, causing network traffic to continue to rise, forming a normal high-traffic network environment, which has also led to network congestion, bandwidth bottlenecks, and frequent service interruptions. In order to improve this problem, we now use multiple network links to share traffic in a high-traffic network environment, effectively improving the quality of network services.

[0003] Since the bandwidth of a single network link is limited, when using multiple network links to share traffic, there may be traffic exceeding the limit. In this case, link switching is required to achieve traffic migration. For example, the invention patent with the authorization announcement number CN108965123B records a link switching method and a network communication system, which uses a monitoring host device to monitor the connectivity status of the first link between the first network device and the second network device; when the connectivity status of the first link is detected to be abnormal, the configuration data packet of the second link is sent to the first network device, and the first network device can configure itself locally through the configuration data packet, so that the second link can be enabled, thereby achieving link switching, so that the link can be switched when the administrator cannot handle it in time, avoiding long-term business interruption. Although this solution realizes the automatic switching of network links in a high-traffic network environment, it lacks attention to whether the network link operation risk supports switching during switching. In the pursuit of switching efficiency, it is easy to perform one-time switching, which may cause data packet loss or incorrect transmission during switching, affecting data integrity and business operations, and may also cause the link to be switched to carry too much traffic in a short time, causing service interruption, affecting business continuity, and is not conducive to the effective implementation of network link switching.

[0004] In addition, the above solution does not take single-link migration and multi-link migration into consideration when performing traffic migration based on network link switching requirements, which can easily lead to the solidification of the migration mode and the hidden danger of incompatibility with the switching requirements. This will not only affect the efficiency of switching implementation, but also may cause the previous migration operations to become invalid due to the need to change the migration mode when identifying that the switching is blocked during the switching process, resulting in a waste of resources and time costs. Summary of the invention

[0005] In view of the shortcomings of the prior art mentioned above, the purpose of the present invention is to provide a variety of network information management systems based on artificial intelligence, which can simultaneously monitor the traffic and operational risks when switching network links in a high-traffic network environment, and extract the network link traffic distribution characteristics at the same time, and then combine the two to perform targeted selection and dynamic switching of network link switching modes, thereby effectively making up for the problems mentioned in the background technology.

[0006] The purpose of the present invention can be achieved through the following technical solutions: A multi-network information management system based on artificial intelligence, including: a network link operation monitoring module, which is used to perform traffic monitoring and operation risk monitoring on multiple network links existing in a high-traffic network environment.

[0007] The traffic migration demand identification module is used to identify the migration demand based on the traffic monitoring results of each network link, and select the network links with traffic migration demand as the excess links, and record the migration time of the demand.

[0008] The healthy link screening module is used to retrieve the traffic and operation risk monitoring results of other network links except the links exceeding the standard at the time of demand migration, thereby screening out healthy links.

[0009] The adaptation migration mode determination module is used to analyze the traffic distribution characteristics of the excess link at the time of demand migration, where the traffic distribution characteristics include the demand migration traffic and the traffic presentation properties, and simultaneously analyze the traffic migration support capacity of the healthy link, thereby determining the adaptation migration mode, and selecting the target link when the adaptation migration mode is determined to be single-link migration, and selecting several adaptive healthy links when the adaptation migration mode is determined to be multi-link migration.

[0010] The single-link gradual migration module is used to gradually migrate traffic based on the real-time monitoring results of the operating risks of the excess link at the time of required migration when the determined adaptation migration mode is single-link migration.

[0011] The traffic allocation dynamic adjustment module is used to allocate initial migration traffic to several adapted healthy links when the determined adapted migration mode is multi-link migration, and dynamically adjust traffic allocation during the migration process.

[0012] Compared with the prior art, the beneficial effects of the present invention are as follows: (1) The present invention monitors the traffic and operation risks during the operation of multiple network links, and identifies the over-standard links and healthy links based on the monitoring results. At the same time, the adaptive migration mode is selected by analyzing the traffic distribution characteristics of the over-standard links and the traffic migration support capacity of the healthy links. This reflects the targeted and adaptive migration of traffic in network link switching, and can maximize the dual guarantee of switching effect and saving switching resources on the basis of improving switching efficiency.

[0013] (2) After selecting the adaptive migration mode, the present invention gradually and dynamically migrates the traffic based on the traffic and operation risk monitoring results of the link associated with the switching, thereby effectively avoiding the one-time switching of the link under the operation risk state, and to a certain extent reducing the incidence of incorrect transmission and service interruption during the switching process, greatly improving the integrity and continuity of data transmission during the switching process, and facilitating the effective and reliable implementation of network link switching. BRIEF DESCRIPTION OF THE DRAWINGS

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

[0015] Figure 1 It is a schematic diagram of the connection of each module of the system of the present invention.

[0016] Figure 2 Schematic diagram of the network link monitoring architecture in the present invention.

[0017] Figure 3 It is a schematic diagram of the process of gradual migration of traffic under single link migration in the present invention. DETAILED DESCRIPTION

[0018] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0019] Reference Figure 1 As shown, the present invention proposes a variety of network information management systems based on artificial intelligence, including a network link operation monitoring module, a traffic migration demand identification module connected to the network link operation monitoring module and a healthy link screening module, a healthy link screening module connected to the traffic migration demand identification module, an adaptation migration mode determination module connected to both the traffic migration demand identification module and the healthy link screening module, a single link gradual migration module and a traffic distribution dynamic adjustment module connected to the adaptation migration mode determination module.

[0020] It should be added that the above system also uses monitoring equipment, processors and control centers in its execution. The monitoring equipment is used to monitor the traffic and operation risks of network links. The processor is used to identify over-standard links and healthy links based on the monitoring results of network link traffic and operation risks. The control center is used to specifically control the traffic migration of over-standard links. The connection architecture of the monitoring equipment, processor and control center in the system is shown in Figure 2 shown.

[0021] The network link operation monitoring module is used to perform traffic monitoring and operation risk monitoring on multiple network links existing in a high-traffic network environment.

[0022] In the specific implementation of the above scheme, the traffic monitoring of the network link can be carried out by hardware devices or software devices. In a further example, the hardware device can be a network probe or a traffic collector, and the software device can be a traffic analysis tool or a performance monitoring tool. These devices monitor the network link traffic mostly by capturing the network data packets of the network link, and then parsing the captured data packets to extract the traffic data.

[0023] In another specific implementation of the above scheme, the operation risk monitoring of the network link is essentially the monitoring of performance indicators such as network link delay, jitter, packet loss rate, etc., so a network performance monitor can be used for operation risk monitoring.

[0024] Applied to the above solution, the operation risk monitoring is implemented as follows: the delay duration, packet loss rate, and jitter duration are monitored in real time during the operation of each network link, and the expression Calculate the operation risk of each network link, where , , They represent the packet loss rate, delay duration, and jitter duration of the network link respectively. , They represent the ideal delay time and the ideal jitter time respectively. represents a set constant, and , for example, .

[0025] It should be added that the ideal delay duration and ideal jitter duration of a network link are related to the bandwidth configuration of the network link, and can be obtained from the configuration description based on the bandwidth of the network link. Furthermore, the ideal delay duration and ideal jitter duration are used in the calculation of the operation risk of the network link in order to eliminate the dimension.

[0026] It should be emphasized that the reason why delay duration, packet loss rate, and jitter duration are used as monitoring data in the operation risk monitoring of network links is that these indicators can provide a comprehensive view of the health status of the network, which is specifically reflected in: Delay duration is used to measure the time it takes for a data packet to travel from the source to the destination. High latency may affect the real-time performance of network applications, so monitoring latency can help discover and solve problems with slow response speeds. Packet loss rate refers to the proportion of data packets lost during transmission. A high packet loss rate usually means that the network link is unreliable, which may be caused by link congestion, equipment failure, or network interference, affecting data integrity and application performance. Jitter duration is the fluctuation of measured latency. High jitter may lead to unstable data transmission. These indicators have a great impact on the healthy operation of the network. By monitoring these indicators, comprehensive, intuitive, and quantitative data on the health status of the network can be provided, providing a basic guarantee for traffic migration.

[0027] The traffic migration demand identification module is used to identify the migration demand based on the traffic monitoring results of each network link, and select the network link with traffic migration demand as the excess link, and record the demand migration time.

[0028] As an embodiment of the above scheme, the migration demand is identified by the following process: the traffic data of each network link captured in real time is compared with the traffic threshold of the corresponding network link. If at a certain moment the migration demand analysis of a certain network link is higher than the traffic threshold of the network link, it is identified that there is a traffic migration demand for the network link at that moment.

[0029] It should be noted that the flow threshold setting of a network link is usually determined by the bandwidth of the network link. This is because the bandwidth is the maximum data transmission rate that the network link can carry. The flow threshold needs to be set within a proportional range of the bandwidth to ensure that the network link is not overloaded. Generally, 70% of the link bandwidth is set as the flow threshold. For example, the bandwidth of the network link is 1 Gbps, and the flow threshold can be set to 700 Mbps.

[0030] The healthy link screening module is used to retrieve the traffic and operation risk monitoring results of other network links except the excessive links at the time of demand migration, thereby screening out healthy links. The specific operation is: compare the traffic data of other network links except the excessive links at the time of demand migration with the traffic threshold of the corresponding network links to calculate the operation idleness of other network links. The specific calculation formula is: , where Indicates the traffic data of other network links at the time of demand migration. Indicates the traffic threshold of other network links.

[0031] The operating idleness and operating risk of other network links are compared with the effective operating idleness and allowed operating risk set by the system. For example, the effective operating idleness can be 0.7 and the allowed operating risk can be 0.3. The effective operating idleness and allowed operating risk are set to assist in the screening of healthy links. If the operating idleness of a network link is higher than or equal to the effective operating idleness and the operating risk is less than or equal to the allowed operating risk, then the network link is regarded as a healthy link.

[0032] The present invention takes into account both idle traffic and operational risk when screening healthy links, and can provide links with sufficient idleness and high operational stability for subsequent traffic migration of excessive links, thereby greatly improving the guarantee of successful and effective traffic migration.

[0033] The adaptation migration mode determination module is used to analyze the traffic distribution characteristics of the excess link at the time of demand migration, where the traffic distribution characteristics include the demand migration traffic and the traffic presentation properties, and simultaneously analyze the traffic migration support capacity of the healthy link, thereby determining the adaptation migration mode, and selecting the target link when the adaptation migration mode is determined to be a single-link migration, and selecting several adaptation healthy links when the adaptation migration mode is determined to be a multi-link migration.

[0034] As a preferred implementation of the above scheme, the traffic distribution characteristics of the over-standard link at the time of demand migration are analyzed as follows: the traffic data of the over-standard link in the historical operating period is retrieved, and the time period regularity analysis is performed to identify the traffic distribution period of the over-standard link and the traffic data of the corresponding period, where the traffic distribution period includes the peak traffic period, the average traffic period and the low traffic period.

[0035] It should be added that the historical operation time period mentioned above should be limited to the current time to avoid the traffic data retrieved being too old and unable to be close to the actual situation, thereby affecting the accuracy and reliability of the subsequent time period regularity analysis. Therefore, by limiting the historical operation time period to the current date, it helps to ensure the timeliness and applicability of the analysis. In addition, the range of the historical operation time period should not be too short or too long. If it is too short, the amount of traffic data retrieved will be small and difficult to be representative. If it is too long, it will increase the time of the time period regularity analysis and reduce the analysis efficiency. Exemplarily, the historical operation time period can be limited to 3 months or half a year.

[0036] In the specific implementation of the above scheme, firstly, the historical flow data of the over-standard link is retrieved from the network monitoring log to ensure that the data includes the timestamp and flow data; then the retrieved historical flow data is cleaned to remove outliers and irrelevant data to ensure data quality; then the flow data is summarized by time period (such as hours) to calculate the total flow or average flow for each time period; then the flow data for each time period is averaged to find out the time period when the flow data is significantly higher than the average level, determine the peak flow period, find out the time period when the flow data is significantly lower than the average level, determine the trough flow period, find out the time period when the flow data is at the average level, determine the average flow period, thereby classifying the time period into peak flow period, average flow period and trough flow period; finally, the flow data for each classified time period is calculated.

[0037] It needs to be explained that network links are affected by different online activities of users in different time periods during operation. For example, the difference in activity during working hours, rest time and peak hours makes its traffic distribution show a regular time distribution.

[0038] The traffic distribution period to which the demand migration moment belongs and the traffic data of the corresponding period are obtained based on the demand migration moment, and the traffic data of the traffic distribution period corresponding to the demand migration moment is used as the demand migration traffic.

[0039] It should be noted that when migrating traffic on the over-standard links, the traffic data of the over-standard links at the time of demand migration is not used as the demand migration traffic. Instead, it is taken into account that traffic migration is not completed in an instant but is a process. Therefore, by analyzing the regularity of the traffic distribution of the over-standard links in the time period, the traffic data of the corresponding traffic distribution period at the time of demand migration is used as the demand migration traffic. It can more accurately reflect the actual migration demand, and then more reasonably select the healthy links to be migrated, avoiding insufficient links due to demand fluctuations during traffic migration. At the same time, it can also better prevent link overload that may occur during the migration process, making the traffic migration process smoother.

[0040] The monitoring time series traffic data within the set time period before the demand migration time is retrieved with the demand migration time as the limit to perform traffic fluctuation analysis. For example, the time period is set to 10 minutes, and the analysis result is compared with the preset permission threshold. For example, the permission threshold of traffic fluctuation is 0.4. If the traffic fluctuation analysis result is higher than the permission threshold, the traffic is identified as short-term burst traffic. Otherwise, the traffic is identified as long-term load.

[0041] In an optional method, the traffic fluctuation analysis can extract the monitored time series traffic data within the set period before the demand migration moment to perform maximum value calculation to obtain the maximum flow and minimum flow, and then divide the maximum flow by the minimum flow after taking the difference between the maximum flow and the minimum flow to obtain the traffic fluctuation result.

[0042] It is important to know that the reason for determining the nature of traffic (short-term burst traffic or long-term load) based on traffic fluctuations is that the fluctuation characteristics of traffic can reveal the actual usage pattern and demand of traffic. Specifically, short-term burst traffic usually manifests as fast and significant traffic peaks and large traffic fluctuations, while long-term loads usually manifest as steady and continuous traffic growth or stable loads.

[0043] The present invention can provide reference guidance for the formation of excessive links by determining whether the nature of the traffic is short-term burst traffic or long-term load based on the analysis results of the traffic fluctuations of the excessive links, and thus can select the traffic migration mode more accurately and reasonably.

[0044] Further preferably, the flow migration support capacity of the healthy link is analyzed as follows: the flow data of the healthy link at the time of demand migration is subtracted from the flow threshold of the corresponding network link to obtain the idle flow of the healthy link.

[0045] The traffic migration support capacity of the healthy link is obtained by subtracting the idle traffic of the healthy link from the demand migration traffic of the exceeded link and then comparing it with the idle traffic of the healthy link. When the traffic migration support capacity sign of a healthy link is positive, it indicates that the idle traffic of the healthy link can meet the demand migration traffic. When the traffic migration support capacity sign of a healthy link is negative, it indicates that the idle traffic of the healthy link cannot meet the demand migration traffic.

[0046] Further preferably, the adaptive migration mode is determined by referring to the following process: the traffic migration support of the healthy link is combined with the traffic presentation property of the excess link into the determination model , get the adaptation migration mode , in the model Indicates the traffic migration support of a healthy link. Indicates the traffic characteristics of the link that exceeds the limit. Indicates existence, Indicates any Indicates and.

[0047] It is important to understand that when the traffic on the overloaded link is short-term burst traffic, single-link migration is preferred. This is because short-term burst traffic is a traffic peak that occurs in a short period of time and usually returns to normal levels soon after the event. Therefore, the main challenge of this type of traffic is to handle the traffic surge in a short period of time. Large-scale traffic migration may not be necessary. At the same time, compared with multi-link migration, single-link migration is simpler to operate and less complex to implement and monitor. For short-term traffic surges, this simplified operation method can reduce management overhead and potential implementation risks. Therefore, choosing single-link migration can quickly and effectively respond to traffic surges while maintaining network flexibility and stability.

[0048] When the traffic on the overloaded link is characterized by long-term load, multi-link migration is preferred. The reason is that for long-term load, the network needs to maintain stability and high performance under continuous high traffic conditions. A single link may not be able to meet the needs of long-term high load, resulting in performance degradation and potential network bottlenecks. Multi-link migration can ensure network stability and high performance by sharing the load. In addition, a single link may become a single point of failure in the network under long-term high load conditions. If there is a problem with the link, it may have a serious impact on the entire network. Multi-link migration can reduce this risk and enhance network reliability through redundant links.

[0049] The present invention monitors traffic and operation risks during the operation of multiple network links, identifies over-standard links and healthy links based on the monitoring results, and selects an adaptive migration mode by analyzing the traffic distribution characteristics of over-standard links and the traffic migration support of healthy links. This embodies the targeted and adaptive migration of traffic in network link switching, and can maximize the dual guarantees of switching effect and saving switching resources on the basis of improving switching efficiency, avoid frequent or unnecessary link switching, reduce network overhead and service interruptions, and not only improve the stability and reliability of the network, but also improve user experience and enhance the flexibility and response capabilities of network management.

[0050] Further preferably, when determining that the adaptive migration mode is single-link migration, alternative healthy links with a traffic migration support greater than 0 are screened out from the healthy links, and the alternative healthy links corresponding to the minimum operating risk are extracted as the target links. When determining that the adaptive migration mode is multi-link migration, since the healthy links are already links with an operating risk less than the allowed operating risk during screening, the selection is made based on the traffic migration support, and the healthy links are specifically arranged in order of traffic migration support from large to small, and then a number of healthy links that meet the needs are extracted from the arrangement results as adaptive healthy links.

[0051] The single-link gradual migration module is used to gradually migrate traffic based on the real-time monitoring results of the operating risk of the excess link at the time of required migration when the determined adaptation migration mode is single-link migration, see Figure 3 As shown, the specific implementation is as follows: the operation risk of the target link at the time of demand migration is calculated inversely, normalized and taken as a percentage to obtain the initial migration ratio of the excess link to the target link at the time of demand migration.

[0052] It should be noted that the inverse calculation of the operation risk in the above-mentioned process results in the operation safety situation. Since the value of the operation risk is not 0, the inverse calculation of the operation risk is meaningful. The operation safety situation is then normalized to limit its value to between 0 and 1 to facilitate the calculation of the initial migration ratio.

[0053] The initial migration flow from the excess link to the target link is obtained by multiplying the required migration flow of the excess link by the initial migration ratio.

[0054] During the migration process, the operation risk and flow of the link exceeding the standard are monitored in real time, and the real-time operation risk and real-time flow of the link exceeding the standard after the migration are obtained.

[0055] Migration stages are set, wherein the migration duration can be estimated through historical migration records, and then the estimated migration duration is divided into migration stages. Specifically, the migration period can be divided into 2-minute intervals at the beginning of the migration, and the traffic threshold of the link that exceeds the standard in the current migration stage is compared with the traffic difference of the real-time traffic, and the operation risk is formed into a traffic difference and operation risk fluctuation curve in a coordinate system with migration time as the horizontal axis and traffic difference and operation risk as the vertical axis. The changing trend of the curve is extracted to identify idle traffic and controllable risks. At the same time, the ratio of the migrated traffic in the current migration period to the required migration traffic is obtained as the migrated rate.

[0056] In the specific application of the above scheme, the extraction of the change trends of the flow difference fluctuation curve and the operation risk fluctuation curve can be achieved by performing differential or derivative calculations on the fluctuation curves to obtain the change rate of the fluctuation curve as the change trend.

[0057] In further applications, the specific operations for identifying idle traffic and controllable risks based on the changing trend of the curve are as follows: extracting the changing trend sign of the traffic difference fluctuation curve. If it is positive, it means that the traffic difference between the traffic threshold of the exceeded link and the real-time traffic is gradually increasing, which means that the real-time traffic is gradually decreasing; extracting the changing trend sign of the operation risk fluctuation curve. If the sign is negative, it means that the operation risk of the exceeded link is showing a downward trend. Therefore, when the changing trend sign of the traffic difference fluctuation curve is positive and the changing trend sign of the operation risk fluctuation curve is negative, it is identified that the traffic and risk of the exceeded link are controllable; otherwise, it is identified that the traffic or risk of the exceeded link is uncontrollable.

[0058] The migrated rate of the current migration period is compared with the limit value set by the system, where the limit value set by the system is 80%. If the migrated rate of the current migration period reaches the limit value, it indicates that most of the traffic has been redistributed to the new link. The reversal may cause traffic redistribution, interfere with the already stable traffic balance, and affect the overall performance of the network. At this time, the migration ratio is continued to be calculated based on the operation risk of the target link, and then the traffic is migrated from the excess link to the target link. On the contrary, if the migrated rate of the current migration period is less than the limit value, it means that the traffic migration is still in progress. If it is identified that the traffic and risk of the link exceeding the limit in the current migration period are controllable, the migrated traffic is reversed from the target link, which can avoid further investment in the new link or adjustment of resources, and use the existing stable link to process the traffic. At the same time, business continuity can be ensured and service interruption can be prevented. In addition, since the single-link migration traffic reversal is also convenient to operate, the traffic reversal can be performed at a lower operating cost. If it is identified that the traffic or risk of the link exceeding the limit in the current migration period is uncontrollable, the migration traffic is continued to be calculated based on the operation risk of the target link, and then the traffic is migrated from the excess link to the target link.

[0059] The traffic allocation dynamic adjustment module is used to allocate initial migration traffic to a number of adapted healthy links when the determined adapted migration mode is multi-link migration, and dynamically adjust traffic allocation during the migration process.

[0060] In the specific implementation, the initial migration traffic is allocated to several adaptive healthy links as follows: the operation risk of each adaptive healthy link at the time of demand migration is combined with the traffic migration support capacity into the formula Count the distribution value of each adaptive health link, where Indicates the traffic migration support capacity of the adapted healthy link. It represents the operational risk of adapting to healthy links. It can be seen from the formula that the greater the traffic migration support, the smaller the operational risk and the greater the allocation value.

[0061] The allocation value of each adapted healthy link is proportionally calculated to obtain the initial migration allocation ratio of each adapted healthy link.

[0062] The initial allocated migration flow of each adapted healthy link is obtained by multiplying the required migration flow of the excess link by the initial migration allocation ratio of each adapted healthy link.

[0063] More specifically, the dynamic adjustment of traffic allocation during the migration process refers to the following process: after the migration is performed according to the initial allocation migration traffic of each adapted healthy link, the operation risk and traffic migration support capacity of each adapted healthy link are monitored in real time.

[0064] Similarly, the operation risk and traffic migration support capacity of each adaptive healthy link during the migration process are allocated and valued, and the migration process allocation ratio of each adaptive healthy link is obtained based on the allocated value.

[0065] The real-time migrated traffic of the link that exceeds the limit is counted, and the required migration traffic of the link that exceeds the limit is subtracted from the real-time migrated traffic to obtain the remaining migration traffic.

[0066] The remaining migration traffic is combined with the migration process allocation ratio of each adapted healthy link to obtain the process allocation migration traffic of each adapted healthy link.

[0067] It should be understood that when multi-link migration is selected, traffic will not be rotated even if the excess link traffic and operational risks are controllable. On the one hand, multi-link migration aims to optimize network load and distribute traffic to multiple links. Rotation may disrupt this load balance and affect overall performance and stability. On the other hand, multi-link migration can make network resources more efficiently utilized. Rotation may cause some link resources to be idle, thereby wasting resources. On the other hand, traffic rotation under multi-link migration is complex and has high operating costs.

[0068] After selecting the adaptive migration mode, the present invention gradually and dynamically migrates the traffic according to the traffic and operation risk monitoring results of the link associated with the switching, thereby effectively avoiding the one-time switching of the link under the operation risk state, and to a certain extent reducing the occurrence rate of incorrect transmission and service interruption during the switching process, greatly improving the integrity and continuity of data transmission during the switching process, and facilitating the effective and reliable implementation of network link switching.

[0069] The above contents are merely examples and explanations of the concept of the present invention. Those skilled in the art may make various modifications or additions to the specific embodiments described or replace them in a similar manner. As long as they do not deviate from the concept of the invention or exceed the scope defined by the present invention, they shall all fall within the protection scope of the present invention.

Claims

1. A multi-network information management system based on artificial intelligence, characterized in that: include: The network link operation monitoring module is used to monitor the traffic and operation risks of multiple network links in a high-traffic network environment; A traffic migration demand identification module is used to identify migration demands based on traffic monitoring results of each network link, select network links with traffic migration demands as excess links, and record the migration demand time; The healthy link screening module is used to retrieve the traffic and operation risk monitoring results of other network links except the links exceeding the standard at the time of demand migration, thereby screening out healthy links; The module for determining the adaptive migration mode is used to analyze the traffic distribution characteristics of the over-standard link at the time of demand migration, where the traffic distribution characteristics include the demand migration traffic and the traffic presentation properties, and simultaneously analyze the traffic migration support of the healthy link, thereby determining the adaptive migration mode, and selecting the target link when the adaptive migration mode is determined to be single-link migration, and selecting several adaptive healthy links when the adaptive migration mode is determined to be multi-link migration; A single-link gradual migration module is used to gradually migrate traffic based on the real-time monitoring results of the operating risk of the excess link at the time of required migration when the determined adaptation migration mode is single-link migration; A traffic allocation dynamic adjustment module, used to allocate initial migration traffic to a number of adapted healthy links when the determined adapted migration mode is multi-link migration, and dynamically adjust traffic allocation during the migration process; The traffic distribution characteristics of the over-limit link at the time of demand migration are analyzed as follows: Retrieve the traffic data of the link that exceeds the standard during the historical operation period, and perform a regularity analysis of the period, thereby identifying the traffic distribution period of the link that exceeds the standard and the traffic data of the corresponding period, where the traffic distribution period includes the traffic peak period, the traffic average period, and the traffic valley period; Based on the demand migration time, the traffic distribution period to which the demand migration time belongs and the traffic data of the corresponding period are obtained, and the traffic data of the traffic distribution period corresponding to the demand migration time is used as the demand migration traffic; The monitoring time series traffic data within the set period before the demand migration time is retrieved to perform traffic fluctuation analysis, and the analysis result is compared with the preset permission threshold. If the traffic fluctuation analysis result is higher than the permission threshold, the traffic is identified as short-term burst traffic, otherwise the traffic is identified as long-term load. The process of analyzing the traffic migration support of healthy links is as follows: The idle flow of the healthy link is obtained by subtracting the network data packet of the healthy link at the time of demand migration from the flow threshold of the corresponding network link; The traffic migration support capacity of the healthy link is obtained by calculating the difference between the idle traffic of the healthy link and the demand migration traffic of the excess link and comparing it with the idle traffic of the healthy link. The adaptation migration mode is determined by referring to the following process: Substitute the traffic migration support capacity of the healthy link and the traffic presentation properties of the excessive link into the determination model , get the adaptation migration mode , in the model Indicates the traffic migration support of a healthy link. Indicates the traffic characteristics of the link that exceeds the limit. Indicates existence, Indicates any Indicates and.

2. The artificial intelligence-based multi-network information management system according to claim 1, characterized in that: The operational risk monitoring is implemented as follows: During the operation of each network link, the delay duration, packet loss rate, and jitter duration are monitored in real time, and the expression Calculate the operation risk of each network link, where , , They represent the packet loss rate, delay duration, and jitter duration of the network link respectively. , They represent the ideal delay time and the ideal jitter time respectively. represents a set constant, and .

3. The artificial intelligence-based multi-network information management system according to claim 2, characterized in that: The migration requirement identification process is as follows: The traffic information of each network link captured in real time is compared with the traffic threshold of the corresponding network link. If the migration demand analysis of a network link at a certain moment is higher than the traffic threshold of the network link, it is identified that there is a traffic migration demand for the network link at that moment.

4. The artificial intelligence-based multi-network information management system according to claim 3, characterized in that: The health link screening process is as follows: Compare the flow information of other network links except the link that exceeds the limit at the time of demand migration with the flow threshold of the corresponding network link to calculate the operation idleness of other network links; The operating idleness and operating risk of other network links are compared with the effective operating idleness and allowed operating risk set by the system. If the operating idleness of a network link is higher than or equal to the effective operating idleness and the operating risk is less than or equal to the allowed operating risk, then the network link is regarded as a healthy link.

5. The artificial intelligence-based multi-network information management system according to claim 2, characterized in that: The traffic migration is implemented gradually as follows: The operation risk of the target link at the time of demand migration is calculated inversely and then normalized to obtain the initial migration ratio of the excess link to the target link at the time of demand migration; The initial migration flow of the excess link to the target link is obtained by multiplying the required migration flow of the excess link by the initial migration ratio; During the migration process, the operation risk and traffic of the link exceeding the standard are monitored in real time, and the real-time operation risk and real-time traffic of the link exceeding the standard after the migration are obtained; Set the migration stage, and form a flow difference and operation risk fluctuation curve by taking the flow threshold of the link exceeding the standard in the current migration stage and the flow difference of the real-time flow and the operation risk in a coordinate system with the migration time as the horizontal axis and the flow difference and operation risk as the vertical axis. Extract the change trend of the curve to identify the controllability of flow and risk. At the same time, obtain the ratio of the migrated flow in the current migration period to the required migration flow as the migrated rate. The migrated rate of the current migration period is compared with the boundary value set by the system. If the migrated rate of the current migration period reaches the boundary value, the migration ratio calculation is continued based on the operation risk of the target link, and then the traffic is migrated from the excess link to the target link. Conversely, if the migrated rate of the current migration period is less than the boundary value, if it is identified that the traffic and risk of the excess link in the current migration period are controllable, the migrated traffic is returned from the target link. If it is identified that the traffic or risk of the excess link in the current migration period is uncontrollable, the migration traffic calculation is continued based on the operation risk of the target link, and then the traffic is migrated from the excess link to the target link.

6. The artificial intelligence-based multi-network information management system according to claim 2, characterized in that: The allocation of initial migration traffic to a number of adapted healthy links is implemented as follows: The operational risk of each adaptive healthy link at the time of demand migration is combined with the traffic migration support to import the formula Count the distribution value of each adaptive health link, where Indicates the traffic migration support capacity of the adapted healthy link. Indicates the operational risk of the adapted health link; Performing proportional calculation on the allocation value of each adaptive healthy link to obtain the initial migration allocation ratio of each adaptive healthy link; The initial allocated migration flow of each adapted healthy link is obtained by multiplying the required migration flow of the excess link by the initial migration allocation ratio of each adapted healthy link.

7. The artificial intelligence-based multi-network information management system according to claim 6, characterized in that: The process of dynamically adjusting traffic distribution during the migration process is as follows: After migrating the traffic according to the initial allocation of each adapted healthy link, monitor the operation risk and traffic migration support capacity of each adapted healthy link in real time; Similarly, the operation risk and traffic migration support of each healthy link during the migration process are calculated according to the allocation value, and the migration process allocation ratio of each healthy link is obtained based on the allocation value; Count the real-time migrated traffic of the link that exceeds the limit, and subtract the required migration traffic of the link that exceeds the limit from the real-time migrated traffic to obtain the remaining migration traffic; The remaining migration traffic is combined with the migration process allocation ratio of each adapted healthy link to obtain the process allocation migration traffic of each adapted healthy link.