IPv6 network data flow monitoring method

Abstract

The application discloses an IPv6 network data flow monitoring method, which integrates network layer and application layer probing. First, the evaluation index and weight are determined, covering ping, route tracing, first screen time, download speed, etc., and the corresponding weight is allocated according to the influence degree of each index on network quality. Second, the expected standard threshold Ti and the reference value Bi are set for each index to quantify the network quality. The probing data is collected from multiple IPv6 operators and terminal devices in the network and sent back to the server. The server compares the actual measurement value with the set value, calculates the single item score, and then weighted sum to get the comprehensive score. Finally, according to the evaluation standard and grade division, the detailed probing report and network problem analysis report are generated, and the IPv6 network quality is evaluated comprehensively and accurately. This method realizes the comprehensive and accurate evaluation of IPv6 network quality by selecting the probing index, setting the standard threshold, and monitoring multiple operators and devices, and provides strong support for network optimization and upgrading.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of IPv6 data monitoring, and particularly relates to an IPv6 network data flow monitoring method. BACKGROUND

[0002] With the widespread deployment and application of IPv6 (Internet Protocol version 6) network technology, its network quality monitoring and evaluation is particularly important. As a new generation of network protocol, IPv6 network aims to solve the problem of IPv4 address space exhaustion and provides higher security and better performance. However, the quality of IPv6 network, including connection stability, response speed, data transmission efficiency, etc., is directly related to the network experience of users and the effect of various applications.

[0003] Traditionally, network quality monitoring mainly focuses on the network layer, such as evaluating the basic connectivity, delay and packet loss of the network through ping test, route tracing, etc. However, these methods can only provide basic performance information at the network layer and cannot fully reflect the performance of the network in supporting specific applications. With the increasing richness and diversity of network applications, users have higher requirements for network quality, and relying solely on network layer monitoring cannot meet the actual needs.

[0004] Therefore, application layer stress testing evaluation has gradually attracted attention. Application layer stress testing simulates user behavior, such as accessing web pages, playing videos, downloading files, etc., to evaluate the performance of the network in supporting these specific applications. This type of stress testing can provide data closer to the actual experience of users, such as first screen time, download speed, and streaming smoothness, thus more comprehensively reflecting network quality. SUMMARY

[0005] The present application provides an IPv6 network data flow monitoring method, which provides an evaluation method combining network layer and application layer stress testing to more comprehensively and accurately evaluate the quality of IPv6 network by reasonably selecting stress testing indicators and assigning weights, setting expected standards and thresholds, and monitoring on multiple operators and terminal devices, etc.

[0006] The present application provides an IPv6 network data flow monitoring method, comprising:

[0007] S101, defining network layer and application layer stress testing task evaluation indicators and weights;

[0008] S102, setting network quality expected standard threshold Ti and reference value Bi for each indicator;

[0009] S103, collecting stress testing data sent after responding to the stress testing task through multiple in-network terminal devices of multiple IPv6 network operators.

[0010] S104, send the dial test data back to the server, and the server receives the dial test data from the terminal device on the network;

[0011] S105, for each index, compare the actual measurement value Mi with the threshold value Ti and the benchmark value Bi, and calculate the single-index score;

[0012] S106, the single-index scores of each index are weighted and summed to obtain a comprehensive score;

[0013] S107, set the evaluation criteria and grade division, and generate a dial test report and a network problem analysis report.

[0014] Preferably, the S101, the network layer and application layer dial test task includes:

[0015] Network latency, packet loss rate, latency, route hop count, first screen time, total download time, download speed, first play duration, page element total loading time weight, buffering time weight; In addition, the build chain time, DNS time, download availability, number of stalls, bit rate, play availability, page availability; In actual application, according to the importance of each index on network quality, adjust the corresponding weight W1, W2,..., Wn, ensure that the sum of all weights is 1.

[0016] Preferably, the S102, the specific network quality expectation standard threshold value Ti and the benchmark value Bi setting includes:

[0017] The expectation standard of network latency, packet loss rate, build chain time, DNS time, number of stalls, first screen time, total download time, first play duration, page element total loading time, and buffering time is a reverse index;

[0018] The expectation standard of download speed, bit rate, download availability, play availability, and page availability is a positive index; The threshold value Ti is the maximum or minimum value of the index data that can be accepted, and the benchmark value Bi is the ideal value that is expected to be reached.

[0019] Preferably, the network quality expectation standard threshold value Ti and the benchmark value Bi setting further includes:

[0020] For the expectation standard being a reverse index representing a high network quality index, the single-index score formula is as follows:

[0021]

[0022] For the expectation standard being a positive index representing a high network quality index, the single-index score formula is as follows:

[0023]

[0024] Ti is the threshold value, Mi is the actual measured value, and Bi is the reference value or the worst possible value.

[0025] Preferably, the S105, the comprehensive score includes:

[0026] Set a weight Wi for each index, and the sum of the weights should be 1; Set a threshold value Ti for each index, and calculate the single item score Si according to the comparison between the actual measured value and the threshold value, and finally weight the sum of the single item scores to obtain the comprehensive score CS; wherein the comprehensive score formula is as follows:

[0027]

[0028] Preferably, the S102 includes:

[0029] S201, dividing a specific period into multiple unit time periods, and setting a load threshold value for each time period;

[0030] S202, at the beginning of each unit time period, monitoring the network load in real time, and adjusting the dialing test task according to the preset load threshold value;

[0031] S203, in each unit time period, collecting dialing test data obtained after executing the dialing test task, and performing preliminary analysis.

[0032] Preferably, the S201 includes:

[0033] S301, continuously monitoring the change of network load in the initial period;

[0034] S302, if the load changes significantly in a short time, the unit time period is considered to be adjusted;

[0035] S303, setting the load change rate rule for period adjustment;

[0036] S304, monitoring the load of the network in real time in the dynamically set unit time period;

[0037] S305, setting the priority of the task according to the complexity of the dialing test task and the influence degree on the network.

[0038] Preferably, the S303 includes:

[0039] When the load change rate exceeds a certain threshold value, the current unit time period is shortened to half of the original; when the load is relatively stable, the unit time period is extended;

[0040] For most network environments, the load change rate threshold is set between 10% and 20%, when the load change rate exceeds 20%, the current unit period is shortened to half of the original; when the load change rate is less than 5%, the unit period is extended.

[0041] Preferably, the S302, adjusting the unit period comprises:

[0042] S401, set two overall completion time thresholds Tmax, namely a fixed sequence task total completion time threshold Tmax1 and a dynamic sequence task total completion time Tmax2;

[0043] S402, for each subtask in the fixed sequence task and the dynamic sequence task, set an individual completion time threshold respectively;

[0044] S403, when executing the fixed sequence, the fixed dial test task sequence is fixed, and the start time, end time and execution sequence of each subtask are recorded in detail;

[0045] S404, when executing the dynamic sequence task, the dial test task is randomly allocated, and the start time, end time and execution sequence of each subtask are recorded in detail;

[0046] S405, according to the recorded subtask time, the total completion time of the entire fixed sequence task and the overall completion time of the entire dynamic sequence task are calculated;

[0047] S406, compare the calculated fixed sequence task total completion time and dynamic sequence task total completion time with Tmax1 and Tmax2 set respectively; if the total completion time is less than or equal to the corresponding Tmax value, it is considered that the network performance is up to standard; otherwise, it is considered that the network performance is not up to standard.

[0048] Preferably, the S401 comprises:

[0049] Tmax1, Tmax2 threshold will be used as the basis for evaluating whether the network performance is up to standard, and Tmax1, Tmax2 threshold is set according to historical network state stable data.

[0050] One or more technical solutions provided in the present application have at least the following technical effects or advantages:

[0051] 1、A series of representative and comprehensive indicators are selected, which not only cover the basic performance parameters of the network layer such as network delay, packet loss rate, etc., but also involve key experience indicators of the application layer such as first screen time, download speed, etc. Such index design can comprehensively reflect the overall quality of IPv6 network, ensuring the comprehensiveness and accuracy of the evaluation results. Secondly, by setting clear weights for each indicator, the system can scientifically and reasonably evaluate the importance of different indicators on network quality, so as to obtain more valuable comprehensive score. This kind of summary evaluation method helps network operators quickly understand the overall performance of the network, and provides strong data support for subsequent optimization and upgrading.

[0052] 2、A comparison method based on actual measurement value, threshold value and reference value is used to objectively and accurately evaluate the actual performance level of each indicator. By setting clear threshold and reference values, the system can judge whether each indicator meets the standard and calculate the single score accordingly. This scoring method not only has clear evaluation criteria, but also can dynamically adjust the score according to the change of actual measurement value, ensuring the real-time and accuracy of the evaluation results. In addition, the single indicator score can provide reliable basic data for the calculation of comprehensive score, further improving the reliability and reference value of the evaluation results.

[0053] 3、Join the unit period analysis network quality and adjust the load threshold of the test, this scheme brings significant advantages for network operation and maintenance. By dividing the specific period into multiple unit periods and conducting in-depth analysis of network quality for each period, network operation and maintenance personnel can more accurately master the performance of the network in different time periods. This fine monitoring method helps to discover and solve potential problems in the network in a timely manner, ensuring the stability and reliability of the network.

[0054] 4、Dynamic setting of unit period significantly improves the flexibility of network quality monitoring. Traditional monitoring methods often use fixed period setting, which cannot adapt to the rapid changes of network load in time. By dynamically setting the unit period, the monitoring strategy can be flexibly adjusted according to the actual situation of network load. In this way, when the network load changes significantly, the monitoring period can be quickly adjusted to more accurately reflect the network status and provide stronger support for network operation and maintenance.

[0055] 5. By measuring the total time required to process predefined, fixed-sequence dialing tasks that must be completed in sequence, the network's response speed and processing capacity can be intuitively assessed. When executing fixed-sequence tasks, detailed records of the start and end times and execution order of each subtask help identify network performance issues. By analyzing subtask completion times and inter-task switching times, network performance issues can be identified and targeted optimization can be implemented. Furthermore, a set threshold value, Tmax1, for the total completion time of fixed-sequence tasks provides a benchmark for network performance. In practice, current network performance can be compared with this benchmark to determine whether network performance meets the standards and provide guidance for performance optimization.

[0056] 6. Dynamically sequenced tasks can dynamically adjust task order or complete multiple tasks simultaneously based on network conditions, helping to test the network's flexibility and ability to handle complex, integrated tasks. By randomly assigning test tasks and recording their execution in detail, we can gain insight into the network's adaptability and stability in the face of ever-changing task demands. During the execution of dynamically sequenced tasks, the random nature of task order and execution time can expose potential network issues, such as resource contention and deadlocks. This provides an opportunity to promptly identify and resolve these issues, thereby improving network reliability and stability. Furthermore, the set threshold Tmax2 for the total completion time of dynamically sequenced tasks provides a clear evaluation metric for the network's overall performance. By comparing the actual completion time with the Tmax2 value, we can accurately determine whether the network's performance in handling complex, integrated tasks meets the target, providing clear guidance for performance optimization. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] Figure 1 The present invention provides a flow chart of an IPv6 network data flow monitoring method. DETAILED DESCRIPTION

[0058] To facilitate understanding of the present invention, the present application will be described more comprehensively below with reference to the relevant drawings; the drawings show preferred embodiments of the present invention, but the present invention can be implemented in many different forms and is not limited to the embodiments described herein; on the contrary, the purpose of providing these embodiments is to enable a more thorough and comprehensive understanding of the disclosed content of the present invention.

[0059] It should be noted that the terms “vertical”, “horizontal”, “up”, “down”, “left”, “right” and similar expressions used in this document are for illustrative purposes only and do not represent the only implementation method.

[0060] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs; the terminology used in the description herein is for describing particular embodiments only and is not intended to be limiting of the application; the use herein of the terms "and / or" includes a combination of one or more of the associated listed items, in any of all combinations.

[0061] Figure 1 is a flowchart of an IPv6 network data flow monitoring method according to an embodiment of the application.

[0062] As shown in Figure 1 , an IPv6 network data flow monitoring method comprises the following steps:

[0063] S101, define network layer and application layer probing task evaluation indexes and weights.

[0064] Wherein, the probing task refers to a series of instructions or requests sent by the server to the terminal device in the network quality monitoring process.

[0065] Wherein, the network layer probing task includes ping and / or route tracing, wherein the ping index data includes packet loss rate and delay, and / or the route tracing index data includes route hop count, network delay and route process;

[0066] The network layer probing index can select network delay weight 0.20, packet loss rate weight 0.15, delay weight 0.10, and route hop count weight 0.05.

[0067] The application layer probing index can select first screen time weight 0.10, total download time weight 0.10, download speed weight 0.10, first play duration weight 0.10, page element total loading time weight 0.05, and buffer time weight 0.05.

[0068] In addition, the following can also be selected: link establishment time (time required to establish a network connection), DNS time, download availability, number of stalls, bit rate, play availability, and page availability.

[0069] In actual application, the corresponding weights W1, W2,..., Wn are adjusted and allocated according to the importance of the influence of each index on network quality, to ensure that the sum of all weights is 1.

[0070] S102, set network quality expected standard threshold value Ti and reference value Bi for each index.

[0071] Specifically, the expected standards of network latency, packet loss rate, link-up time, DNS time, number of stalls, first screen time, total download time, first play duration, total page element loading time, and buffer time are reverse indicators, i.e., their values are lower than the expected standards, indicating that the higher the network quality is;

[0072] The expected standards of download speed, bit rate, download availability, play availability, and page availability are positive indicators, i.e., their values are higher than the expected standards, indicating that the higher the network quality is;

[0073] The threshold Ti is the maximum or minimum value of the indicator data that can be accepted (depending on the nature of the indicator), and the benchmark value Bi is the ideal value expected to be achieved, which is used to evaluate the pros and cons of network quality;

[0074] Example threshold values Ping packet loss rate (T1 = 1%, B1 = 0%), Ping latency (T2 = 100ms, B2 = 50ms), route hop count (T3 = 15 hops, B3 = 10 hops), network latency (T4 = 200ms, B4 = 100ms), web page access first screen time (T5 = 5s, B5 = 2s), web page access availability (T6 = 99%, B6 = 100%), total file download time (T7 = 30s, B7 = 15s), file download speed (T8 = 1Mbps, B8 = 5Mbps), video play first play duration (T9 = 2s, B9 = 1s), video play stall number (T10 = 3 times, B10 = 0 times), and specific values need to be set according to actual conditions.

[0075] S103, through a plurality of in-network terminal devices of a plurality of IPv6 network operators, collect ping test data sent after responding to the ping test task.

[0076] Specifically, through a plurality of in-network network terminal devices, by calling an embedded browser kernel to access web pages using an IPv6 network and obtaining web page access indicator data, for example, by calling an embedded player to play online videos using an IPv6 network and obtaining video play indicator data, and by calling an embedded browser kernel to download files using an IPv6 network and obtaining file download indicator data.

[0077] S104, send the ping test data back to the server, and the server receives the ping test data from the in-network terminal device.

[0078] Specifically, by performing the probing task on the network-side device and obtaining the related index data, the index data more truly reflects the actual use of the IPv6 network, and the probing task includes a network layer probing task and an application layer probing task, thereby evaluating the IPv6 network quality from two dimensions of network link and application experience. Moreover, the probing data is obtained by the in-network terminals of multiple operators, thereby realizing the horizontal evaluation of the IPv6 network quality of the multiple operators. In summary, the method provided in the embodiments of the present application realizes the horizontal monitoring of the index data of the network layer and the application layer of the operator IPv6 network, and more truly evaluates the IPv6 network quality.

[0079] S105, for each index, comparing the actual measurement value Mi with the threshold value Ti and the reference value Bi to calculate the single-index score.

[0080] For the expected standards such as packet loss rate and delay, the single-index score formula is as follows:

[0081]

[0082] For the expected standards such as first-screen time and download speed, the single-index score formula is as follows:

[0083]

[0084] Ti is the threshold value, Mi is the actual measurement value, and Bi is the reference value (such as the worst value).

[0085] S106, the single-index scores of the indexes are weighted and summed to obtain a comprehensive score.

[0086] Specifically, a weight (Wi) is set for each index, and the sum of the weights should be 1. Then, a threshold value (Ti) is set for each index, and the single-index score (Si) is calculated according to the comparison between the actual measurement value and the threshold value, and finally the single-index scores are weighted and summed to obtain a comprehensive score (CS).

[0087] The comprehensive score formula is as follows:

[0088]

[0089] S107, setting the evaluation criteria and grade division to generate a probing report and a network problem analysis report.

[0090] The report should include the measurement value, single-index score, comprehensive score and corresponding grade evaluation of each index, and the analysis in the report includes the problems of the network data flow quality, such as a lower score of a certain index, poorer network quality in a certain area, etc.

[0091] The score of the comprehensive score (CS) is the standard corresponding to the score level, which is further limited according to the actual application scene. In the conventional implementation scene, the following score standard and score level can be used to evaluate the network quality:

[0092] Excellent (A) CS >= 90 points, indicating that the network quality is very good, and the network experience is excellent; Good (B) 80 points <= CS < 90 points, indicating that the network quality is good, and the network experience is satisfactory; General (C) 70 points <= CS < 80 points, indicating that the network quality is general, and the network will have slight delay or lag; Poor (D) 60 points <= CS < 70 points, indicating that the network quality is poor, and the network has obvious delay, lag or download failure and other problems; Poor (E) CS < 60 points, indicating that the network quality is very poor, and the network may not be able to normally use network services.

[0093] For example, a large Internet company plans to comprehensively monitor and evaluate the network quality of three large IPv6 network operators (operators A, B and C) it cooperates with, to ensure that the online services (including web browsing, video playback, file download) it provides can run stably and efficiently in the IPv6 environment;

[0094] Network layer probing index weight, threshold and benchmark value: Ping delay weight 0.20, T1 = 100 ms, B1 = 50 ms, Ping packet loss rate weight 0.15, T2 = 1%, B2 = 0%, route tracing-network delay weight 0.10, T3 = 200 ms, B3 = 100 ms, route tracing-route hop count weight 0.05, T4 = 15 hops, B4 = 10 hops;

[0095] Application layer probing index: first screen time weight 0.10, T5 = 5 s, B5 = 2 s, total download time weight 0.10, T6 = 30 s, B6 = 15 s, download speed weight 0.10, T7 = 1 Mbps, B7 = 5 Mbps, first playback time weight 0.10, T8 = 2 s, B8 = 1 s, total page element loading time weight 0.05, T9 = 10 s, B9 = 5 s, buffering time weight 0.05, T10 = 5 s, B10 = 2 s, additional selection link establishment time, DNS time, download availability (weight is allocated to related indexes, total weight is still 1);

[0096] Randomly select 1000 in-network terminal devices (about 333 in each operator) nationwide, ensure uniform distribution of devices, cover different regions and different network environments, and execute network layer and application layer probing tasks to record various index data;

[0097] After performing the ping test task, the terminal device returns the measurement data to the server through the IPv6 network. The server receives and stores the ping test data from each terminal device, preparing for subsequent analysis.

[0098] For a device with a Ping delay measurement value of 60 ms, the score is calculated according to the formula: Similarly, the individual scores of all other indicators are calculated.

[0099] According to the weights and individual scores of each indicator, the comprehensive score of each device is calculated. For example, the comprehensive score of a certain device is: 0.20 x 80 + 0.15 x 95 +... + 0.05 x 70 = 82.5 points.

[0100] The report includes the measurement values of each indicator for each device, individual scores, comprehensive scores, and corresponding grade evaluations. The average comprehensive score of Operator A is 85 points, which is good; the average comprehensive score of Operator B is 80 points, which is general; and the average comprehensive score of Operator C is 75 points, which is poor.

[0101] In the report, the comprehensive score of Operator C is poor, and in the network problem analysis report, it can be seen that the Ping delay and download speed scores are low, which are the main factors affecting the comprehensive score. The network quality in some remote areas is generally poor and needs to be focused on.

[0102] The technical solutions in the above embodiments of the present application have at least the following technical effects or advantages:

[0103] 1. A series of representative and comprehensive indicators are selected. These indicators not only cover basic performance parameters at the network layer, such as network delay and packet loss rate, but also involve key experience indicators at the application layer, such as first screen time and download speed. Such indicator design can comprehensively reflect the overall quality of IPv6 network, ensuring the comprehensiveness and accuracy of the evaluation results. Secondly, by setting clear weights for each indicator, the system can scientifically and reasonably evaluate the importance of different indicators on network quality, thus obtaining a more valuable comprehensive score. This general evaluation method helps network operators quickly understand the overall performance of the network and provides strong data support for subsequent optimization and upgrading.

[0104] 2、Using a comparison method based on actual measurement values, threshold values and benchmark values, this method can objectively and accurately reflect the actual performance level of each indicator. By setting clear threshold values and benchmark values, the system can determine whether each indicator meets the standard and calculate the individual score accordingly. This scoring method not only has clear evaluation criteria, but also dynamically adjusts the score according to the changes in actual measurement values, ensuring the timeliness and accuracy of the evaluation results. In addition, the calculation of individual indicator scores can provide reliable basic data for the derivation of comprehensive scores, further improving the reliability and reference value of the evaluation results.

[0105] Example two: The above example one obtains actual dialing test data by dialing test on the network, compares with the pre-set expected threshold value and benchmark value, and obtains the comprehensive score of the network after adding the weight, ensuring the comprehensiveness and accuracy of the evaluation results. To further realize the network quality evaluation of the network dialing test monitoring method more accurately, the example one is further improved, and the network quality is analyzed in a unit period. The load threshold value of the actual situation in a unit period is used to adjust the dialing test task, so as to ensure that the dialing test process does not cause overload burden to the network, and at the same time meet the network data flow monitoring and specific business needs. On the basis of step S102 of example two, further improvement is made, which is specifically:

[0106] S201, divide a specific period into multiple unit periods, and set a load threshold value for each time period.

[0107] Among them, the unit period is a time period in a day or a specific period, such as an hour, and the load threshold value of each unit period is set according to the business needs. The load threshold value is used to judge the numerical limit of network load situation. Setting the load threshold value can provide basis for subsequent adjustment of dialing test task, and ensure that the dialing test process does not cause too much pressure to the network.

[0108] S202, at the beginning of each unit time period, real-time monitor the network load situation, and adjust the dialing test task according to the pre-set load threshold value.

[0109] Specifically, the time of the unit period and the monitored network are determined, and the open source monitoring tool Nagios or the commercial SolarWinds is selected according to the monitoring target. The selection of the monitoring tool should support the required monitoring indicators and be easy to integrate into the existing IT management system.

[0110] Set the monitoring parameters in the monitoring tool, including monitoring frequency, threshold value, etc. If the monitoring tool needs an agent to collect data, deploy the agent program on the corresponding network.

[0111] For example, when the load is high (>80%), all ping test tasks are stopped to avoid additional pressure on the network; when the load is moderate (60%-80%), simple ping test tasks such as routing path, delay, etc. are performed to evaluate the basic performance of the network; when the load is low (<60%), complex ping test tasks such as video playback, file download, etc. are performed to comprehensively evaluate the network quality.

[0112] S203, in each unit period, collect the ping test data obtained after performing the ping test task and perform preliminary analysis.

[0113] For example, a company A needs to monitor the network load in real time and dynamically adjust the ping test task according to the load to accurately evaluate the network quality. A day is selected as a specific period, a week is divided into 7 days, and each day is divided into 24 unit periods, each of which is 1 hour.

[0114] The load threshold is set to 75% during peak hours (9:00-18:00) on weekdays (Monday to Friday) and 50% during off-peak hours (other times), and the load threshold is set to 60% throughout the day on weekends (Saturday and Sunday).

[0115] Select the open source monitoring tool Nagios, the monitoring frequency is every 5 minutes, configure according to the load threshold set in S201, and deploy Nagios agent program on core routers, switches and servers to collect monitoring data;

[0116] The collected data is analyzed to evaluate the performance of the network under different loads and time periods. When the network load exceeds the set threshold, the ping test task is automatically reduced or stopped to avoid additional pressure on the network; when the network load is lower than the set threshold, the ping test task is automatically increased to more comprehensively evaluate the network quality.

[0117] The technical solutions in the above embodiments of the present application have at least the following technical effects or advantages:

[0118] Joining the unit period analysis of network quality and adjusting the load threshold of ping test brings significant advantages to network operation. By carefully dividing a specific period into multiple unit periods and conducting in-depth analysis of network quality for each period, network operation personnel can more accurately master the performance of the network at different times. This fine monitoring method helps to discover and solve potential problems in the network in a timely manner, thereby ensuring the stability and reliability of the network.

[0119] According to the real-time monitoring of network load, the load threshold of the dial test is adjusted, further optimizing the execution of the dial test task. In the low peak period, by increasing the dial test task, the network performance can be more comprehensively evaluated, and potential problems can be found in advance; while in the peak period, by reducing or stopping the dial test task, the additional pressure on the network caused by the dial test is avoided, ensuring the rational use of network resources.

[0120] Embodiment three: The above embodiment two adds unit period analysis of network quality and adjustment of dial test load threshold not only improves the efficiency and accuracy of network operation and maintenance, but also reduces the operation and maintenance cost, fine management and dynamic adjustment, so that the network can better adapt to the demand of different time periods, ensuring the stable operation and efficient use of the network. In order to more flexibly adapt to the change of network load and more accurately evaluate the network quality, the method of dynamically setting unit period can be used, and the execution of dial test task is dynamically adjusted combined with network load. Based on the step S201 of embodiment two, further improvement is made, which is as follows:

[0121] S301, continuously monitor the change of network load in the initial period.

[0122] S302, if the load changes significantly in a short time, consider adjusting the unit period.

[0123] Specifically, assume a time window of 5 minutes, calculate the load change rate in this time window. If the change rate exceeds the preset threshold, it is considered that the load has changed significantly in a short time, and the unit period needs to be adjusted.

[0124] S303, dynamically set the load change rate rule of period adjustment.

[0125] Specifically, the feedback situation of network dial test data is detected in real time, when the load change rate exceeds a certain threshold, the current unit period is shortened to half of the original; when the load is relatively stable, the unit period can be appropriately extended; if the adjusted unit period again exceeds a certain threshold or is relatively stable, adjust again until the dial test task is completed;

[0126] For most network environments, the load change rate threshold can be set between 10% and 20%, when the load change rate exceeds 20%, the current unit period is shortened to half of the original; when the load change rate is less than 5%, the unit period is extended. These thresholds are set according to actual demand and network conditions.

[0127] The period adjustment rule should aim to balance the fineness of monitoring and operation and maintenance cost. When the load change rate exceeds a threshold (e.g., 15%), the current unit period is shortened to half of the original. For example, if the current period is 1 hour, it is adjusted to 30 minutes. In this way, the load change can be responded to more quickly, and more fine-grained monitoring can be provided;

[0128] When the load change rate is below a certain stable threshold (e.g., 5%) and lasts for a period of time (e.g., two consecutive time windows), the current unit period can be appropriately extended. The extension ratio can be determined according to the actual situation, but it is generally recommended not to exceed 50% of the original period. For example, if the current period is 30 minutes, it can be extended to a maximum of 45 minutes;

[0129] If the load change rate fluctuates within a certain range (e.g., 5% to 15%) and there is no upward or downward trend, the current unit period can be kept unchanged.

[0130] The load change rate formula is as follows:

[0131]

[0132] S304, real-time monitoring of the load of the network in the dynamically set unit period.

[0133] For example, in each dynamically set unit period, the network monitoring tool is used to continue real-time monitoring of the network load.

[0134] S305, setting the priority of the task according to the complexity of the task and the degree of influence on the network.

[0135] For example, route path tracking and delay measurement can be set as high-priority tasks, and video playback testing can be set as medium-priority tasks. The specific priority order needs to be limited in combination with specific application scenarios.

[0136] When the load is high, high-priority tasks are executed first, and medium- and low-priority tasks are suspended or reduced; when the load is moderate, all priority tasks are balanced to ensure that the network quality can be fully evaluated; when the load is low, all tasks can be executed to obtain more comprehensive network quality data.

[0137] The technical solutions in the embodiments of the present application have at least the following technical effects or advantages:

[0138] The dynamic setting of the unit period significantly improves the flexibility of network quality monitoring. Traditional monitoring methods often use fixed period setting and cannot timely adapt to the rapid changes of network load. By dynamically adjusting the unit period, the monitoring strategy can be flexibly adjusted according to the actual situation of network load. In this way, when the network load changes significantly, the monitoring period can be quickly adjusted to more accurately reflect the network status and provide stronger support for network operation and maintenance.

[0139] The task priority is set according to the complexity of the dialing test task and the degree of influence on the network, further optimizing the strategy of network quality monitoring. By reasonably setting the task priority, the dialing test task can be executed under different network load conditions, thereby ensuring the comprehensiveness and accuracy of network quality evaluation.

[0140] Embodiment four: the above embodiment three dynamically sets the unit period, significantly enhancing the flexibility of network quality monitoring, and reasonably sets the task priority according to the complexity of the dialing test task and its influence on the network, thereby further optimizing the overall strategy of network quality monitoring. On this basis, the dynamic time unit task is divided into two types: one is a dialing test task that must be completed in a fixed order, measuring the total time required for the network to complete these tasks to evaluate the response speed and processing capacity of the network; the other is a dialing test task that can be dynamically adjusted in order or completed simultaneously according to the network itself, testing the flexibility of the network and the ability to handle complex and comprehensive tasks. Based on step S302 of embodiment three, further improvements are made, specifically:

[0141] S401, set two overall completion time thresholds Tmax, i.e. fixed sequence task total completion time threshold Tmax1 and dynamic sequence task total completion time Tmax2.

[0142] Specifically, Tmax1 and Tmax2 thresholds will be used as the basis for evaluating whether the network performance meets the standard, and Tmax1 and Tmax2 thresholds are set based on historical network state stable data.

[0143] S402, for each subtask in the fixed sequence task and the dynamic sequence task, set a separate completion time threshold.

[0144] Among them, the threshold of the fixed sequence task and the dynamic sequence task should be set based on the complexity of the subtask under historical data analysis, the historical execution time and the availability of network resources.

[0145] S403, when executing the fixed sequence, the fixed dialing test task sequence is recorded in detail, including the start time, end time and execution order of each subtask.

[0146] S404, when performing the dynamic sequential task, randomly allocate the dialing test task, and record the start time, end time and execution order of each subtask in detail.

[0147] The dynamic sequential task does not consider priority and execution order, and randomly matches the dialing test task. The dialing test task data can be allocated by using a random number generator. For specific steps, please refer to the prior art, which will not be described here.

[0148] S405, according to the recorded subtask time, calculate the total completion time of the entire fixed sequential task and the total completion time of the entire dynamic sequential task.

[0149] The total completion time Tmax is the sum of the completion time of each subtask, plus the necessary task switching time based on actual measurement.

[0150] S406, compare the calculated total completion time of the fixed sequential task and the total completion time of the dynamic sequential task with the set Tmax1 and Tmax2 respectively. If the total completion time is less than or equal to the corresponding Tmax value, it is considered that the network performance meets the standard; otherwise, it is considered that the network performance does not meet the standard.

[0151] For example, first set the total completion time threshold, the fixed sequential task total completion time threshold Tmax1 is 60 seconds, and the dynamic sequential task total completion time threshold Tmax2 is 90 seconds;

[0152] Set the task completion time threshold, the completion time threshold of subtasks A, B and C in the fixed sequential task is 10 seconds, 20 seconds and 30 seconds respectively; the completion time threshold of subtasks D, E and F in the dynamic sequential task is 12 seconds, 18 seconds and 25 seconds respectively, wherein the dialing test task priority and weight of subtasks D, E and F are different;

[0153] For the fixed sequential task, execute according to the predefined order A->B->C, and record the start time, end time and execution order of each subtask in detail;

[0154] For the dynamic sequential task, randomly allocate the dialing test task D->F->E or E->D->F, disturb the order combination, and record the start time, end time and execution order of each subtask in detail;

[0155] The total completion time Tmax1 of the fixed sequential task is the sum of the completion time of subtasks A, B and C, plus the task switching time, wherein the task switching time is assumed to be 5 seconds;

[0156] The total completion time Tmax2 of the dynamic sequential task = the completion time of subtask A + the completion time of subtask B + the completion time of subtask C + the task switching time, wherein the task switching time is assumed to be 5 seconds, but the actual switching time may vary due to different execution sequences of the tasks;

[0157] If Tmax1 is less than or equal to 60 seconds, it is considered that the network performance of the fixed sequential task meets the standard; otherwise, it does not meet the standard.

[0158] If Tmax2 is less than or equal to 90 seconds, it is considered that the network performance of the dynamic sequential task meets the standard; otherwise, it does not meet the standard.

[0159] The technical solutions in the embodiments of the present application have at least the following technical effects or advantages:

[0160] By measuring the total time required to process the fixed sequential test task that is predefined and must be completed in sequence, the response speed and processing capacity of the network can be intuitively evaluated. When executing the fixed sequential task, the start time, end time and execution sequence of each subtask are recorded in detail, which helps to find network performance problems. By analyzing the completion time of the subtasks and the switching time between tasks, the cause of the network performance problem can be identified, so that targeted optimization can be performed. In addition, the set fixed sequential task total completion time threshold Tmax1 provides a benchmark for network performance. In actual application, the performance of the current network can be compared with this benchmark to determine whether the network performance meets the standard, and guidance for performance optimization is provided.

[0161] The dynamic sequential task can dynamically adjust the task sequence or complete multiple tasks at the same time according to the network itself, which greatly helps to test the flexibility of the network and the ability to process complex and comprehensive tasks. By randomly assigning test tasks and recording their execution in detail, the adaptability and stability of the network when facing changing task demands can be understood in depth. During the execution of the dynamic sequential task, due to the randomness of the task sequence and execution time, some potential problems in the network, such as resource competition and deadlock, may be exposed. This provides an opportunity to discover and solve these problems in a timely manner, thereby improving the reliability and stability of the network. In addition, the set dynamic sequential task total completion time threshold Tmax2 provides a clear evaluation standard for the comprehensive performance of the network. By comparing the actual completion time with the value of Tmax2, it can be accurately determined whether the performance of the network in processing complex and comprehensive tasks meets the standard, and a clear direction for performance optimization is provided.

[0162] The above merely describes the preferred embodiments of the present application and is not used to limit the present application. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. A method for monitoring IPv6 network data flow, characterized in that: include: Define the evaluation indicators and weights of network layer and application layer dial testing tasks; Set the network quality expectation threshold Ti and benchmark value Bi for each indicator; Divide a specific cycle into multiple unit time periods and set a load threshold for each time period. During the initial time period, continuously monitor changes in network load. If the load changes significantly within a short period of time, consider adjusting the unit time period. The load change rate rules for periodic adjustments are set. Real-time monitoring is performed on the feedback of network dial-up test data. When the load change rate exceeds a certain threshold, the current unit time period is shortened to half of the original. When the load stabilizes, the unit time period is extended. If the adjusted unit time period again exceeds a certain threshold or becomes relatively stable, it is adjusted again until the dial-up test task is completed. Within dynamically set unit time periods, the network load is monitored in real time. Task priorities are set based on the complexity of the dialing test tasks and their impact on the network. At the beginning of each unit time period, the network load is monitored in real time and dialing test tasks are adjusted based on preset load thresholds. Two overall completion time thresholds, Tmax, are set: Tmax1 for the total completion time of fixed-sequence tasks and Tmax2 for the total completion time of dynamic-sequence tasks. Separate completion time thresholds are set for each subtask within the fixed-sequence and dynamic-sequence tasks. The calculated total completion time of the fixed-sequence and dynamic-sequence tasks is compared with the set Tmax1 and Tmax2, respectively. The total completion time, Tmax, is the sum of the completion times of each subtask, plus any inter-task switching time required based on actual measurements. Fixed-sequence tasks are executed in a predefined order, and the start and end times and execution order of each subtask are recorded in detail. For dynamic-sequence tasks, dialing test tasks are randomly assigned and shuffled, and the start and end times and execution order of each subtask are recorded in detail. Within each unit time period, dialing test data obtained after executing the dialing test tasks is collected and preliminarily analyzed. Collect the dial test data sent after responding to the dial test task through multiple online terminal devices of multiple IPv6 network operators; Send the dial test data back to the server, and the server receives the dial test data from the terminal device on the network; For each indicator, the score of the single indicator is calculated by comparing the actual measured value Mi with the threshold Ti and the benchmark value Bi; The weighted sum of the individual scores of each indicator is used to obtain a comprehensive score; Set assessment criteria and grading, and generate dial test reports and network problem analysis reports.

2. The method for monitoring IPv6 network data flow according to claim 1, wherein: The network layer and application layer dial testing tasks include: Network latency, packet loss rate, delay, routing hops, first screen time, total download time, download speed, first play time, total page element loading time weight, buffering time; Other options include link building time, DNS time, download availability, number of freezes, bit rate, playback availability, and page availability; In practical applications, it is necessary to adjust and assign corresponding weights W1, W2, ..., Wn according to the importance of each indicator to the network quality, so as to ensure that the sum of all weights is 1.

3. The method for monitoring IPv6 network data flow according to claim 1, wherein: The specific network quality expectation standard threshold Ti and reference value Bi are set as follows: The expected standards for network latency, packet loss rate, link establishment time, DNS time, number of freezes, first screen time, total download time, first play time, total page element loading time, and buffering time are reverse indicators; Expected standards for download speed, bitrate, download availability, playback availability, and page availability are positive indicators; The threshold Ti is the maximum or minimum value acceptable to the indicator data. If this value is exceeded, the network quality is considered to be substandard. The benchmark value Bi is the ideal value expected to be achieved.

4. The method for monitoring IPv6 network data flow according to claim 3, wherein: The network quality expectation standard threshold Ti and reference value Bi are set, and further include: For the expected standard, the inverse indicator represents a high network quality indicator, and the individual scoring formula is as follows: For indicators where the expected standard is a positive indicator representing high network quality, the individual scoring formula is as follows: Ti is the threshold value, Mi is the actual measured value, and Bi is the reference value or the worst possible value.

5. The method for monitoring IPv6 network data flow according to claim 1, wherein: The comprehensive score includes: Set a weight Wi for each indicator, and the sum of the weights should be 1; set a threshold Ti for each indicator, and calculate the individual score Si based on the comparison between the actual measurement value and the threshold. Finally, the weighted sum of these individual scores is used to obtain the comprehensive score CS. The comprehensive score formula is as follows: 。 6. The method for monitoring IPv6 network data flow according to claim 1, wherein: The load change rate rules include: When the load change rate exceeds a certain threshold, the current unit period is shortened to half of the original period; when the load is relatively stable, the unit period is extended; For most network environments, the load change rate threshold is set between 10% and 20%. When the load change rate exceeds 20%, the current unit period is shortened to half of the original period; when the load change rate is less than 5%, the unit period is extended.

7. The method for monitoring IPv6 network data flow according to claim 1, wherein: The adjustment unit period includes: Set two overall completion time thresholds Tmax, namely the total completion time threshold Tmax1 for fixed-sequence tasks and the total completion time threshold Tmax2 for dynamic-sequence tasks; For each subtask in the fixed-order task and the dynamic-order task, a separate completion time threshold is set; When executing a fixed sequence, fix the dialing task sequence and record the start time, end time and execution sequence of each subtask in detail; When executing dynamic sequence tasks, randomly assign dialing tasks and record the start time, end time and execution order of each subtask in detail; Based on the recorded subtask times, calculate the total completion time of the entire fixed sequence task and the overall completion time of the entire dynamic sequence task; The calculated total completion time of fixed-sequence tasks and the total completion time of dynamic-sequence tasks are compared with the set Tmax1 and Tmax2 respectively; if the total completion time is less than or equal to the corresponding Tmax value, the network performance is considered to meet the standard; otherwise, the network performance is considered to fail to meet the standard.

8. The method for monitoring IPv6 network data flow according to claim 7, wherein: The method further comprises: The Tmax1 and Tmax2 thresholds will serve as benchmarks for evaluating whether network performance meets standards.

Images