A Fault Diagnosis Method for Computer Network

By deploying multiple distributed fault diagnosis nodes in the computer network, combining the hierarchical topology and traffic distribution of the network, comprehensive real-time monitoring of network status is achieved, and the problems of slow fault diagnosis and poor accuracy in the existing technology are solved, the efficiency and accuracy of fault detection and positioning are improved, and the stability and continuity of the network are ensured.

CN119071140BActive Publication Date: 2025-05-06NANTONG TIANXING INFORMATION TECH CO LTD

Patent Information

Application Number
CN202411480310.8
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-10-23
Publication Date
2025-05-06
Estimated Expiration
2044-10-23

AI Technical Summary

Technical Problem

Existing computer network fault diagnosis methods have performance bottlenecks, single-point failure risks and limitations when facing large-scale networks and complex failures, making it difficult to quickly and accurately detect and locate local faults.

Method used

By deploying multiple distributed fault diagnosis nodes in a computer network, combining the hierarchical topology and traffic distribution of the network, comprehensive real-time monitoring of network status is achieved. Each node independently handles the fault diagnosis task and exchanges information through the communication network. It uses advanced fault diagnosis algorithms and standardized fault information formats to coordinate the fault location and cause, and generate diagnostic results.

Benefits of technology

It realizes comprehensive real-time monitoring of network status, shortens the time for fault detection and location, improves the real-time and accuracy of diagnosis, reduces the time of impact of faults on the network, and ensures the stability and continuity of the network through a fault tolerance mechanism.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119071140B_ABST
    Figure CN119071140B_ABST
Patent Text Reader

Abstract

The present invention discloses a fault diagnosis method for a computer network, and relates to the technical field of computer networks. The specific steps of the diagnosis method are as follows: in a computer network, multiple distributed fault diagnosis nodes are deployed and combined with the hierarchical topological structure and traffic distribution of the network. By deploying multiple distributed fault diagnosis nodes in a computer network, the present invention enables each node to independently process the fault diagnosis task it is responsible for, and at the same time, nodes at different positions communicate with a fault diagnosis center node to realize information sharing. When an abnormality occurs in the network, each node can respond quickly and determine the fault location and cause through collaborative work. Combined with threshold comparison and statistical methods of abnormality detection, the fault source can be accurately located, which greatly improves the accuracy of fault diagnosis and reduces the impact time of the fault on the network.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of computer networks, and in particular to a fault diagnosis method for a computer network. Background Art

[0002] With the rapid development of information technology, computer networks have become an indispensable infrastructure in modern society, supporting the operation and development of all walks of life. However, with the expansion of network scale and the increase of complexity, the frequency and impact range of network failures have also increased. Network failures will not only lead to data loss and service interruption, but may also cause security incidents, causing significant losses to enterprises and individuals. Therefore, how to quickly and accurately diagnose and repair network failures has become an urgent problem to be solved in computer network management.

[0003] At present, traditional centralized fault diagnosis methods often rely on a central node for fault diagnosis when dealing with large-scale network failures, resulting in performance bottleneck problems. When the network scale is large and the number of failures is large, the central node will not be able to respond in time due to limited processing capacity, thereby prolonging the time for fault diagnosis and repair. In addition, centralized methods have the risk of single point failure, that is, if the central node fails, the entire fault diagnosis method will not work properly, which will make the network more vulnerable when a failure occurs, and there are limitations in obtaining network status information. Since the central node can only obtain limited global information, it is difficult to accurately detect and locate local faults in the network.

[0004] To sum up, the existing computer network fault diagnosis methods have obvious shortcomings when facing large-scale networks and complex faults. In order to solve these problems, how to deploy multiple distributed fault diagnosis nodes in the network to achieve comprehensive monitoring of the network status and rapid network fault diagnosis methods has become an urgent solution to ensure the stable operation of computer networks. Summary of the invention

[0005] The purpose of the present invention is to make up for the shortcomings of the prior art and to provide a computer network fault diagnosis method, which can achieve comprehensive real-time monitoring of the network status by deploying multiple distributed fault diagnosis nodes and combining the hierarchical topology structure and traffic distribution of the network, effectively shorten the time of fault detection and positioning, and improve the real-time nature of diagnosis. At the same time, by using advanced fault diagnosis algorithms and standardized fault information formats, the accuracy and consistency of the diagnosis results are ensured, providing a reliable basis for subsequent fault repair.

[0006] In order to solve the above technical problems, the present invention provides the following technical solutions: a computer network fault diagnosis method, the diagnosis method comprising the following steps:

[0007] S100, in a computer network, multiple distributed fault diagnosis nodes are deployed in combination with the hierarchical topology and traffic distribution of the network;

[0008] Fault diagnosis nodes are deployed at the core layer, aggregation layer, and access layer to comprehensively monitor the network status at different levels;

[0009] Each of the distributed fault diagnosis nodes can independently process the fault diagnosis task it is responsible for, and the fault diagnosis nodes at different locations communicate with the fault diagnosis center node;

[0010] The fault diagnosis center node is integrated by a monitoring unit, a diagnosis unit, a communication unit, a control unit and a storage unit;

[0011] S200, the fault diagnosis nodes exchange information through a communication network to share fault diagnosis information;

[0012] When a node detects a fault, it sends the fault information to nearby nodes in a standardized format;

[0013] The fault information includes the fault type, occurrence time, impact scope, priority, and associated device ID of the fault;

[0014] S300, when a node detects a network anomaly, it immediately starts a fault diagnosis process and notifies other nodes, wherein the fault diagnosis process includes anomaly detection, fault location, fault repair, and generates a diagnosis result;

[0015] The anomaly detection uses threshold comparison and statistical methods to detect network anomalies; the fault location is based on the anomaly detection information, the information shared by each node and its own diagnosis results, and collaboratively determines the fault location and cause; the fault repair uses the fault diagnosis algorithm to formulate a corresponding fault handling plan according to the severity, impact range and repair priority of the fault, and coordinates each fault diagnosis node to implement it; the diagnosis result records the type, time, repair process and diagnostic information of the fault;

[0016] S400: Taking corresponding fault repair suggestions according to the diagnosis result, wherein the repair suggestions include resetting the device, reconfiguring the network parameters, and replacing the hardware.

[0017] Furthermore, the core layer fault diagnosis nodes are used to monitor the stability of high-bandwidth traffic and the status of key routing nodes; the aggregation layer fault diagnosis nodes are used to monitor the communication quality and traffic balance between different subnets; and the access layer fault diagnosis nodes are used to monitor the connection stability and user experience indicators of terminal devices.

[0018] Furthermore, the monitoring unit in the fault diagnosis center node is used to collect network status data in real time, the diagnosis unit is used to analyze the monitoring data, identify and locate faults, the communication unit is used to exchange information with other nodes, the control unit is used to manage and coordinate the diagnostic activities of the nodes, and the storage unit is used to save fault history data, analysis results and configuration information.

[0019] Furthermore, after receiving the fault information, the nearby nodes perform further analysis and confirmation based on the monitored data. If the same fault is detected on multiple nodes, the central node repairs the nearest fault and transmits the repair result to other faulty nodes. Other nodes perform self-processing based on the repair result of the central node.

[0020] Furthermore, each node can share fault diagnosis results, and share the same type of fault diagnosis results to other nodes through the central node. Each node adds new diagnostic information to the knowledge base and obtains diagnostic records shared by other nodes from the central node.

[0021] Furthermore, the anomaly detection uses threshold comparison and statistical methods to detect anomalies, by collecting network performance indicator data, calculating the network fluctuation range, comparing it with normal network communication fluctuations based on historical data, and setting dynamic thresholds to detect anomalies, that is, for network parameters as variables , the current value is And the value in the time series is , define the normal fluctuation threshold range as ,in and The calculation formula is: , ,but , ,in and is the actual network adjustment coefficient, is the average value of the network parameters, reflecting the central trend of the network parameters. is the standard deviation of the network parameters, which measures the discreteness of the values, that is, the fluctuation of the network parameters. Current value of satisfy or When the value exceeds the normal fluctuation range, a preliminary abnormal judgment is triggered, that is, the current value exceeds the normal fluctuation range and an abnormal situation exists, and statistical methods are used for further confirmation.

[0022] Furthermore, the anomaly detection is based on preliminary anomaly judgment, further confirmed using statistical methods, and defining statistics , and its calculation formula is: , this statistic It measures the ratio of the distance between the current value and the average value to the standard deviation. A large ratio indicates that the current value deviates from the normal situation. The statistical anomaly threshold is set to ,when When the network is in an abnormal state, it is determined that the network is in an abnormal state. For the abnormal state, the fault is located based on the information shared by each node and its own diagnosis results.

[0023] Furthermore, the fault repair method can accurately determine the severity, impact range and repair priority of the fault using the fault diagnosis algorithm, so as to formulate a corresponding fault handling plan, that is, the fault severity parameter , Impact range parameters , Fault repair priority parameters , for the fault severity parameter , and its calculation formula is: , Indicates the degree of abnormal data transmission, which is determined by comparing the data transmission rate difference between the normal situation and the current network, that is, ,in is the data transmission rate under normal circumstances, is the current data transmission rate, Indicates the fault duration, Indicates the degree of network load change, which is determined by the ratio of the difference between the current network load and the average network load to the average network load, that is, ,in is the current network load, is the average network load, , , is the weight coefficient.

[0024] Furthermore, the influence range parameter Considering the number of devices and network areas affected by the fault, as well as the importance weight of each device area and the degree of impact of the fault, the impact range parameter , and its calculation formula is ,in, is the number of devices and network areas affected by the failure, Indicates The importance weights of each device and network area, Indicates The weighting of the degree to which each device and network area is affected by the failure.

[0025] Furthermore, the fault repair priority parameter The calculation formula is: ,in, , , To adjust the coefficient, according to the fault repair priority parameter Determine the fault handling plan and set the priority threshold, that is, the high priority threshold , Medium priority threshold ,for , emergency repair measures are taken, that is, immediately resetting key equipment and adjusting core network parameters. , priority repair measures are taken, that is, gradually troubleshooting faulty nodes and reconfiguring some network parameters. , take routine repair measures, perform maintenance and troubleshooting.

[0026] Compared with the prior art, this computer network fault diagnosis method has the following beneficial effects:

[0027] 1. The present invention deploys multiple distributed fault diagnosis nodes in a computer network so that each node can independently handle the fault diagnosis task it is responsible for. At the same time, nodes at different locations communicate with the fault diagnosis center node to achieve information sharing. When an abnormality occurs in the network, each node can respond quickly and determine the fault location and cause through collaborative work. Combined with the threshold comparison and statistical methods of anomaly detection, the fault source can be accurately located, which greatly improves the accuracy of fault diagnosis and reduces the impact time of the fault on the network.

[0028] 2. The present invention realizes distributed processing of fault diagnosis tasks by deploying multiple fault diagnosis nodes in the network. Even if some nodes fail, other nodes can still continue to perform fault diagnosis, thereby ensuring the overall stability and continuity of the computer network. In addition, the present invention also designs a fault-tolerant mechanism. Even if some nodes fail, the normal operation of fault diagnosis can still be maintained through the cooperation of other normal nodes. This high reliability and fault tolerance enable the distributed fault diagnosis method of the present invention to adapt to various complex network environments, providing a more solid guarantee for the stable operation of the network.

[0029] Other advantages, objectives and features of the present invention will be set forth in part in the following description and, in part, will be apparent to those skilled in the art based on an examination of the following or may be taught from the practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention, and for ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0031] Figure 1 The present invention is an operation flow chart of a computer network fault diagnosis method.

[0032] Figure 2 This is a diagram showing the composition of a fault diagnosis center node unit in a computer network fault diagnosis method. DETAILED DESCRIPTION

[0033] The technical solutions in the embodiments of the present invention are described clearly and completely below. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of them. 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.

[0034] Embodiment 1

[0035] This embodiment provides a specific application process of a computer network fault diagnosis method in a network of a large data center to ensure efficient and stable operation of the data center.

[0036] First, according to the hierarchical network topology and traffic distribution of the data center, distributed fault diagnosis nodes are carefully deployed in the core layer, aggregation layer and access layer. The fault diagnosis nodes deployed in the core layer, such as node A, focus on monitoring the stability of high-bandwidth traffic and the status of key routing nodes. It collects the performance indicators of the core switch in real time, including traffic peak, packet loss rate and other data. By analyzing these data, node A can promptly detect abnormal fluctuations in high-bandwidth traffic and potential problems of key routing nodes. For example, when it is detected that the traffic on a port of the core switch suddenly increases and exceeds the normal threshold, node A will immediately trigger an abnormal alarm and pass the relevant information to other nodes and the fault diagnosis center node; the fault diagnosis node of the aggregation layer, taking node B as an example, focuses on the communication between different subnets. Quality and traffic balance, it closely monitors the connection between the aggregation switch and each subnet, including indicators such as data packet delay and bandwidth utilization. When abnormalities are found in inter-subnet communication, such as a sudden increase in data packet delay or excessive bandwidth utilization, Node B will quickly analyze and determine which subnet or link has a problem, and feedback the details to the fault diagnosis center node; the fault diagnosis node at the access layer, such as Node C, is mainly responsible for detecting the connection stability and user experience indicators of the terminal device. It monitors the connection between the server and the terminal device in real time, including connection success rate, response time, etc. Once the terminal device connection is unstable, such as frequent connection interruptions or long response time, Node C will immediately start the fault diagnosis process and send the fault information to the nearby Node B in a standardized format.

[0037] Each of these distributed fault diagnosis nodes can independently handle the fault diagnosis tasks it is responsible for. For example, when node C detects that the connection between a server and a terminal device is abnormal, it will immediately start the fault diagnosis process. First, node C will use threshold comparison and statistical methods to detect network anomalies. It collects network performance indicator data related to the connection, such as the current data transmission rate, delay, etc., and compares it with the benchmark data under normal conditions. If the current value exceeds the defined normal fluctuation threshold range, such as the data transmission rate is significantly lower than the normal level, node C will trigger a preliminary abnormality judgment. Then, node C will use statistical methods for further confirmation and calculate the statistic S, the calculation formula of which is: , and compares it with the set statistical abnormality threshold T. If S>T, node C determines that the network is in an abnormal state and sends the fault information to the nearby node B in a standardized format.

[0038] After receiving the fault information, node B will conduct further analysis and confirmation based on the data it has monitored. It will check the connection between the aggregation switch it monitors and the relevant subnets to see if there are any abnormal signs related to the fault reported by node C. At the same time, nodes A, B, and C all communicate with the fault diagnosis center node and upload their respective diagnosis results to the center node.

[0039] The fault diagnosis center node is integrated with a monitoring unit, a diagnostic unit, a communication unit, a control unit and a storage unit. The monitoring unit collects network status data in real time, including diagnostic results uploaded by each node, network performance indicators, etc. The diagnostic unit uses advanced algorithms to analyze these data, identify and locate faults. For example, by comprehensively analyzing the data uploaded by nodes A, B, and C, the diagnostic unit finds that the abnormal connection between the server and the terminal device is due to a failure in a port at the access layer, which may be caused by a hardware failure or a network configuration problem. The communication unit sends the fault location information to the relevant nodes, the control unit coordinates the entire fault repair process, and the storage unit saves the fault history data and analysis results for subsequent data analysis and optimization. For example, the storage unit will record the occurrence time of the fault, the fault type, the ID of the relevant equipment, the data during the diagnosis process, and the final repair plan, etc., to provide a reference for the diagnosis and processing of similar faults in the future.

[0040] In the fault repair stage, a corresponding fault handling plan is formulated according to the severity of the fault, the scope of impact and the repair priority. Since the fault affects the communication between some servers and terminal devices, it is a medium priority fault and priority repair measures are taken. Specifically, the control unit will dispatch technicians to the access layer. The technicians will quickly locate the port where the fault occurs based on the information provided by the fault diagnosis center node, and then reconfigure the network parameters of the relevant ports, such as adjusting the bandwidth limit, modifying the data packet priority, etc., and check the hardware equipment, such as checking whether the physical connection of the port is normal and whether the hardware is damaged. During the whole process, the diagnosis results are recorded in detail, including the type of fault, the time of occurrence, the repair process and the diagnostic information. These records are not only helpful for summarizing and analyzing the fault, but also provide valuable experience for subsequent network maintenance and optimization. By using the computer network fault diagnosis method of the present invention, the network of the large data center can timely and accurately diagnose and repair various faults, ensure the stable operation of the data center, and improve the reliability and service quality of the data center.

[0041] In summary, by applying the computer network fault diagnosis method of the present invention, the network of the large data center can diagnose and repair various faults in a timely and accurate manner, ensuring the stable operation of the data center, and proving its significant advantages in improving the efficiency and accuracy of network fault diagnosis.

[0042] Embodiment 2

[0043] This embodiment provides a specific application process of a computer network fault diagnosis method in an intelligent transportation system.

[0044] First, according to the hierarchical topology and traffic distribution of the intelligent transportation network, distributed fault diagnosis nodes are deployed at each key node. In the core layer, the deployed fault diagnosis nodes, such as node A, focus on monitoring the stability of high-bandwidth data transmission between the traffic control center and each subsystem, as well as the status of key routing nodes. For example, the monitoring unit collects network status data in real time, including parameters such as data transmission rate and delay. For the network parameter P, its current value is And the value in the time series is , define the normal fluctuation threshold range as ,in and The calculation formula is: , ,but , ,in and is the actual network adjustment coefficient, is the average value of the network parameters, reflecting the central trend of the network parameters. is the standard deviation of the network parameters, which measures the discreteness of the values, that is, the fluctuation of the network parameters. Current value of satisfy or When the value exceeds the normal fluctuation range, the initial abnormal judgment is triggered, that is, the current value exceeds the normal fluctuation range, and there is an abnormal situation. The statistical method is further used to confirm and define the statistical quantity. , set the statistical anomaly threshold to T. When S>T, it is determined that the network is in an abnormal state. At the convergence layer, fault diagnosis nodes such as node B focus on the communication quality and traffic balance between different areas. For example, they monitor the information transmission between each intersection to ensure the accurate transmission and coordination of traffic signals. At the access layer, fault diagnosis nodes such as node C detect the connection stability and data transmission quality of terminal devices (such as cameras, sensors, etc.). When node C detects that the camera data transmission at a certain intersection is abnormal, the fault diagnosis process is immediately started. First, perform anomaly detection to determine whether the network is abnormal according to the above formula. Assuming that the current data transmission rate Data transfer rates significantly lower than normal , triggers a preliminary abnormal judgment, which is further confirmed by statistical methods, calculates the value of S, and compares it with T to determine that the network is in an abnormal state. Based on the abnormal detection information, the fault is located. Node C sends the fault information to the nearby node B in a standardized format, and communicates with node A and the fault diagnosis center node at the same time. The diagnostic unit of the fault diagnosis center node collaboratively determines the fault location and cause based on the information shared by each node and its own diagnostic results. Through analysis, it is determined that the network equipment at the intersection may be faulty, or the signal transmission may be unstable due to external interference.

[0045] Then, perform fault repair and use the fault diagnosis algorithm to calculate the fault severity parameter , Impact range parameters , Fault repair priority parameters , assuming α=0.4, β=0.3, γ=0.3, Indicates the degree of abnormal data transmission, which is determined by comparing the data transmission rate difference between the normal situation and the current network, that is, , Indicates the fault duration, Indicates the degree of network load change, which is determined by the ratio of the difference between the current network load and the average network load to the average network load, that is, ,but , influence range parameter , where N is the number of devices and network areas affected by the failure, Indicates The importance weights of each device and network area, Indicates The weight of the degree to which each device and network area is affected by the fault, and the fault repair priority parameter (in , , is the adjustment factor), according to the calculated The value is compared with the set priority threshold, assuming that the high priority threshold 0.8, medium priority threshold is 0.5, if , take emergency repair measures and immediately dispatch technicians to the intersection to reset related equipment and adjust network parameters; if , then take priority repair measures, gradually check the faulty nodes, and reconfigure some network parameters; if , take routine repair measures, perform maintenance and troubleshooting.

[0046] During the entire process, the diagnosis results are recorded in detail, including the type of fault (such as equipment failure, signal interference, etc.), time, repair process and diagnostic information, providing a basis for subsequent network optimization and maintenance.

[0047] In summary, by using the computer network fault diagnosis method of the present invention, the intelligent transportation network system can timely and accurately diagnose and repair various faults to ensure the efficient operation of the transportation system.

[0048] It will be apparent to those skilled in the art that the invention is not limited to the details of the exemplary embodiments described above and that the invention can be implemented in other specific forms without departing from the spirit or essential features of the invention. Therefore, the embodiments should be considered exemplary and non-limiting in all respects, and the scope of the invention is defined by the appended claims rather than the foregoing description, and it is intended that all variations falling within the meaning and scope of the equivalent elements of the claims be included in the invention. Any reference numeral in a claim should not be considered as limiting the claim to which it relates.

Claims

1. A computer network fault diagnosis method, characterized in that: The diagnostic method includes the following steps: S100, in a computer network, multiple distributed fault diagnosis nodes are deployed in combination with the hierarchical topology and traffic distribution of the network; Fault diagnosis nodes are deployed at the core layer, aggregation layer, and access layer to comprehensively monitor the network status at different levels; Each of the distributed fault diagnosis nodes can independently process the fault diagnosis task it is responsible for, and the fault diagnosis nodes at different locations communicate with the fault diagnosis center node; The fault diagnosis center node is integrated with a monitoring unit, a diagnosis unit, a communication unit, a control unit and a storage unit; S200, the fault diagnosis nodes exchange information through a communication network to share fault diagnosis information; When a node detects a fault, it sends the fault information to nearby nodes in a standardized format; The fault information includes the fault type, occurrence time, impact scope, priority, and associated device ID of the fault; S300, when a node detects a network anomaly, it immediately starts a fault diagnosis process and notifies other nodes, wherein the fault diagnosis process includes anomaly detection, fault location, fault repair, and generates a diagnosis result; The anomaly detection uses threshold comparison and statistical methods to detect network anomalies; the fault location is based on the anomaly detection information, the information shared by each node and its own diagnosis results, and collaboratively determines the fault location and cause; the fault repair uses the fault diagnosis algorithm to formulate a corresponding fault handling plan according to the severity, impact range and repair priority of the fault, and coordinates each fault diagnosis node to implement it; the diagnosis result records the type, time, repair process and diagnostic information of the fault; S400: Taking corresponding fault repair suggestions according to the diagnosis result, wherein the repair suggestions include resetting the device, reconfiguring the network parameters, and replacing the hardware.

2. A computer network fault diagnosis method according to claim 1, characterized in that: The fault diagnosis nodes of the core layer are used to monitor the stability of high-bandwidth traffic and the status of key routing nodes; The aggregation layer fault diagnosis node is used to monitor the communication quality and traffic balance between different subnets; the access layer fault diagnosis node is used to monitor the connection stability and user experience indicators of terminal devices.

3. A computer network fault diagnosis method according to claim 1, characterized in that: The monitoring unit in the fault diagnosis center node is used to collect network status data in real time, the diagnostic unit is used to analyze the monitoring data, identify and locate faults, the communication unit is used to exchange information with other nodes, the control unit is used to manage and coordinate the diagnostic activities of the nodes, and the storage unit is used to save fault history data, analysis results and configuration information.

4. A computer network fault diagnosis method according to claim 1, characterized in that: After receiving the fault information, the nearby nodes perform further analysis and confirmation based on the monitored data. If the same fault is detected on multiple nodes, the central node repairs the nearest fault and transmits the repair result to other faulty nodes. Other nodes perform self-processing based on the repair result of the central node.

5. A computer network fault diagnosis method according to claim 4, characterized in that: Each node can share fault diagnosis results and share the same type of fault diagnosis results to other nodes through the central node. Each node adds new diagnostic information to the knowledge base and obtains diagnostic records shared by other nodes from the central node.

6. A computer network fault diagnosis method according to claim 1, characterized in that: The anomaly detection uses threshold comparison and statistical methods to detect anomalies. It collects network performance indicator data, calculates the network fluctuation range, compares it with normal network communication fluctuations based on historical data, and sets dynamic thresholds to detect anomalies. That is, for network parameters as variables , the current value is And the value in the time series is , define the normal fluctuation threshold range as ,in and The calculation formula is: , ,but , ,in and is the actual network adjustment coefficient, is the average value of the network parameters, reflecting the central trend of the network parameters. is the standard deviation of the network parameters, which measures the discreteness of the values, that is, the fluctuation of the network parameters. Current value of satisfy or When the value exceeds the normal fluctuation range, a preliminary abnormal judgment is triggered, that is, the current value exceeds the normal fluctuation range and an abnormal situation exists, and statistical methods are used for further confirmation.

7. A computer network fault diagnosis method according to claim 6, characterized in that: The anomaly detection is based on preliminary anomaly judgment, further confirmed using statistical methods, and defines statistics , and its calculation formula is: , this statistic It measures the ratio of the distance between the current value and the average value to the standard deviation. A large ratio indicates that the current value deviates from the normal situation. The statistical anomaly threshold is set to ,when When the network is in an abnormal state, it is determined that the network is in an abnormal state. For the abnormal state, the fault is located based on the information shared by each node and its own diagnosis results.

8. A computer network fault diagnosis method according to claim 1, characterized in that: Fault repair uses fault diagnosis algorithms to accurately determine the severity, impact range and repair priority of the fault, so as to develop a corresponding fault handling plan, namely the fault severity parameter , Impact range parameters , Fault repair priority parameters , for the fault severity parameter , and its calculation formula is: , Indicates the degree of abnormal data transmission, which is determined by comparing the data transmission rate difference between the normal situation and the current network, that is, ,in is the data transmission rate under normal circumstances, is the current data transmission rate, Indicates the fault duration, Indicates the degree of network load change, which is determined by the ratio of the difference between the current network load and the average network load to the average network load, that is, ,in is the current network load, is the average network load, , , is the weight coefficient.

9. A computer network fault diagnosis method according to claim 8, characterized in that: The impact range parameters Considering the number of devices and network areas affected by the fault, as well as the importance weight of each device area and the degree of impact of the fault, the impact range parameter , and its calculation formula is ,in, is the number of devices and network areas affected by the failure, Indicates The importance weights of each device and network area, Indicates The weighting of the degree to which each device and network area is affected by the failure.

10. A computer network fault diagnosis method according to claim 1, characterized in that: The fault repair priority parameter The calculation formula is: ,in, , , To adjust the coefficient, according to the fault repair priority parameter Determine the fault handling plan and set the priority threshold, that is, the high priority threshold , Medium priority threshold ,for , emergency repair measures are taken, that is, immediately resetting key equipment and adjusting core network parameters. , priority repair measures are taken, that is, gradually troubleshooting faulty nodes and reconfiguring some network parameters. , take routine repair measures, perform maintenance and troubleshooting.

Citation Information

Patent Citations

  • Fault diagnosis method and system for computer network

    CN115580519A

  • Electric power communication fault diagnosis and prevention method and system

    CN117675522A

Cited By

  • Detecting anomalous resource distribution patterns in distributed artificial intelligence-based agent networks

    US12592897B2

  • Detecting anomalous resource distribution patterns in distributed artificial intelligence-based agent networks

    US20260012432A1