Network Fault Root Cause Analysis via Knowledge Graph Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current fault tree methods are inaccurate in locating fault root causes in networks due to their focus on single-device fault propagation, failing to consider inter-device fault propagation in actual networks.
Innovation Solution
A fault root cause determining method and apparatus that utilizes a knowledge graph to identify fault features across the entire network, incorporating fault propagation relationships and inference models trained on known fault scenarios to accurately determine fault causes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If a fault tree method based on single-device fault propagation is used, then the fault analysis process is simple, but the accuracy of fault root cause location deteriorates due to inability to consider inter-device fault propagation
Solution Approach 1:
The patent segments the network into multiple devices and constructs fault trees for each device separately. Each fault tree analyzes fault propagation within that specific device, and then the results are integrated across devices. This segmentation allows the system to maintain simplicity in individual device analysis while achieving comprehensive network-wide fault root cause location accuracy by considering inter-device fault propagation through the integration of multiple segmented fault trees.
2Measurement precision
If a knowledge graph covering the entire network is constructed, then the fault root cause location accuracy is improved by considering fault propagation between devices, but the system complexity increases
Solution Approach 1:
Instead of constructing one large complex knowledge graph for the entire network, the patent segments the network into multiple devices and constructs individual fault trees for each device. This segmentation reduces the complexity of knowledge graph construction by breaking down the large-scale network into manageable device-level units, while still achieving comprehensive fault root cause location accuracy through the integration of results from all device-level fault trees.
Solution Approach 2:
The patent transitions from a device-level fault analysis dimension to a network-wide dimension by integrating fault trees from multiple devices. This dimensional change allows the system to achieve comprehensive fault root cause location accuracy across the entire network while managing complexity through the structured integration of device-level analyses, effectively adding a network dimension without overwhelming system complexity.
3Measurement precision
If fault propagation between different devices is considered, then the fault root cause location accuracy is improved, but the computational complexity and data processing requirements increase
Solution Approach 1:
The patent segments the computational task into device-level fault tree constructions and integrations. By first constructing fault trees for individual devices separately and then integrating the results, the system reduces computational complexity at each step compared to constructing one massive network-wide fault tree. This segmented approach allows fault propagation between devices to be considered accurately while managing computational resources efficiently through divide-and-conquer methodology.
Data Source
AI summary
In a fault root cause determining method, a management device obtains a fault feature of a target network in which a fault occurs. The fault feature is determined based on a knowledge graph of the target network in which the fault occurs. The knowledge graph includes a network entity and an abnormal event entity. The abnormal event entity is connected to an abnormal network entity. The abnormal network entity is a network entity that generates an abnormal event in the target network. The abnormal event entity is configured to indicate the abnormal event generated by the abnormal network entity. A type of the network entity is a network device, an interface, a protocol, or a service. The management device determines a fault root cause of the target network based on the fault feature.