# US 2022/0286348 A1 · Method and device for determining root cause of fault, server and computer-readable medium

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2022/0286348 A1
- Document ID: US20220286348A1
- Status: Published application
- Kind: A1
- Application number: 17/637,888
- Filed: 2020-09-03
- Published: 2022-09-08
- Earliest claimed: 2019-09-04
- Assignees: Zte Corporation
- Applicants: Zte Corporation
- Inventors: Junhua HAN
- Web page: https://patentgrep.com/patent/US20220286348A1

## CPC

- H04L 41/0631: using root cause analysis; using analysis of correlation between notifications, alarms or events based on decision criteria, e.g. hierarchy, tree or time analysis
- H04L 41/065: {involving logical or physical relationship, e.g. grouping and hierarchies}
- H04L 41/0677: Localisation of faults
- H04L 41/0894: Policy-based network configuration management
- H04L 41/145: {involving simulating, designing, planning or modelling of a network}
- H04L 41/16: using machine learning or artificial intelligence
- H04L 41/22: comprising specially adapted graphical user interfaces [GUI]
- H04L 45/28: using route fault recovery
- H04L: TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION (arrangements common to telegraphic and telephonic communication H04M)

## Abstract

The present disclosure provides a method and a device for determining a root cause of fault, a server and a computer-readable medium. The method for determining a root cause of fault includes: determining fault characteristic information when a fault occurs in a service, and determining a fault root cause node according to the fault characteristic information and a preset fault propagation graph, the fault propagation graph is configured to represent directed association relationships among different fault propagation nodes, and directed association relationships among a fault node, the fault propagation nodes and fault root cause nodes, and the nodes in the fault propagation graph are configured to represent the fault characteristic information.

## Claims

### Claim 1

A method for determining a root cause of fault, comprising:
determining fault characteristic information in response to a fault occurring in a service; and
determining a fault root cause node according to the fault characteristic information and a preset fault propagation graph, wherein the fault propagation graph is configured to represent directed association relationships among different fault propagation nodes, and directed association relationships among a fault node, the fault propagation nodes and fault root cause nodes, and the nodes in the fault propagation graph are configured to represent the fault characteristic information.

### Claim 2 (depends on 1)

The method of claim 1, wherein after determining the fault root cause node according to the fault characteristic information and the preset fault propagation graph, the method further comprises:
determining a fault root cause probability of the fault root cause node according to the fault root cause node, the fault characteristic information and a preset Bayesian model.

### Claim 3 (depends on 2)

The method of claim 2, wherein determining the fault characteristic information comprises:
determining the service where the fault occurs and a fault type;
determining a path of the service where the fault occurs; and
acquiring an event related to the fault according to the path,
determining the fault root cause node according to the fault characteristic information and the preset fault propagation graph comprises:
selecting a fault propagation graph, a fault node in which is matched with the fault type, from the preset fault propagation graphs; and
determining the fault root cause node according to the selected fault propagation graph, and
determining the fault root cause probability of the fault root cause node according to the fault root cause node, the fault characteristic information and the preset Bayesian model comprises:
selecting a Bayesian model corresponding to the selected fault propagation graph from the preset Bayesian models; and
determining the fault root cause probability of the fault root cause node according to the fault root cause node, the fault type, the event related to the fault and the selected Bayesian model.

### Claim 4 (depends on 3)

The method of claim 3, wherein determining the fault root cause node according to the selected fault propagation graph comprises:
determining a node associated with the fault node in the selected fault propagation graph according to association relationships in the selected fault propagation graph; and
determining a node type of the node; and
determining a suspected fault root cause node at least according to the node type, wherein the suspected fault root cause node is a fault root cause node in the selected fault propagation graph, and the number of suspected fault root cause nodes is less than that of fault root cause nodes in the selected fault propagation graph, and
determining the fault root cause probability of the fault root cause node according to the fault root cause node, the fault type, the event related to the fault and the selected Bayesian model comprises:
determining a fault root cause probability of the suspected fault root cause node according to the suspected fault root cause node, the fault type, the event related to the fault and the selected Bayesian model.

### Claim 5 (depends on 4)

The method of claim 4, wherein determining the suspected fault root cause node at least according to the node type comprises:
in response to the node type of the node being a fault propagation node and a verifiability attribute of the node being “verifiable”, determining whether a rule attribute of the node is satisfied;
in response to the rule attribute being satisfied, determining a next node relative to the node according to the selected fault propagation graph, determining a node type of the next node, and determining the suspected fault root cause node according to the node type of the next node; and in response to the rule attribute being not satisfied, stopping determining the next node relative to the node; and
in response to the node type of the node being the fault propagation node and the verifiability attribute of the node being “unverifiable”, determining the next node relative to the node according to the selected fault propagation graph, determining the node type of the next node, and determining the suspected fault root cause node according to the node type of the node.

### Claim 6 (depends on 1)

The method of claim 1, to further comprising generating the fault propagation graph, which comprises:
acquiring, from a preset database of knowledge and experience of operation and maintenance, fault information comprising fault types, fault root cause types, events related to the fault types, and association relationships among the fault types, the fault root cause types and the events related to the fault types; and
generating a fault propagation graph for each of the fault types according to the fault information, wherein the fault propagation graph is a knowledge graph of fault propagation.

### Claim 7 (depends on 2)

The method claim 2, further comprising building the Bayesian model, which comprises:
deleting a node whose verifiability attribute is “unverifiable” in the fault propagation graph;
deleting association relationships related to the deleted node;
establishing, according to the deleted association relationships, an association relationship between nodes associated with the deleted node, so as to obtain a preprocessed fault propagation graph; and
respectively converting all nodes and association relationships in the preprocessed fault propagation graph into nodes and association relationships in the Bayesian model, so as to build the Bayesian model.

### Claim 8

A device for determining a root cause of fault, comprising: a first determining module and a second determining module, wherein
the first determining module is configured to determine fault characteristic information in response to a fault occurring in a service; and
the second determining module is configured to determine a fault root cause node according to the fault characteristic information and a preset fault propagation graph, wherein the fault propagation graph is configured to represent directed association relationships among different fault propagation nodes, and directed association relationships among a fault node, the fault propagation nodes and fault root cause nodes, and the nodes in the fault propagation graph are configured to represent the fault characteristic information.

### Claim 9 (depends on 8)

The device for determining a root cause of fault of claim 8, further comprising a third determining module configured to determine a fault root cause probability of the fault root cause node according to the fault root cause node, the fault characteristic information, and a preset Bayesian model.

### Claim 10 (depends on 9)

The device for determining a root cause of fault of claim 9, wherein the first determining module is further configured to determine the service where the fault occurs and a fault type, determine a path of the service where the fault occurs, and acquire an event related to the fault according to the path;
the second determining module is further configured to select a fault propagation graph, a fault node in which is matched with the fault type, from the preset fault propagation graphs, and determine the fault root cause node according to the selected fault propagation graph; and
the third determining module is further configured to select a Bayesian model corresponding to the selected fault propagation graph from the preset Bayesian models, and determine the fault root cause probability of the fault root cause node according to the fault root cause node, the fault type, the event related to the fault and the selected Bayesian model.

### Claim 11 (depends on 10)

The device for determining a root cause of fault of claim 10, wherein the second determining module is further configured to determine a node associated with the fault node in the selected fault propagation graph according to association relationships in the selected fault propagation graph, determine a node type of the node, and determine a suspected fault root cause node at least according to the node type, wherein the suspected fault root cause node is a fault root cause node in the selected fault propagation graph, and the number of suspected fault root cause nodes is less than that of fault root cause nodes in the selected fault propagation graph; and
the third determining module is further configured to determine a fault root cause probability of the suspected fault root cause node according to the suspected fault root cause node, the fault type, the event related to the fault and the selected Bayes model.

### Claim 12 (depends on 11)

The device for determining a root cause of fault of claim 11, wherein the second determining module is further configured to determine whether a rule attribute of the node is satisfied in response to the node type of the node being a fault propagation node and a verifiability attribute of the node being “verifiable”;
in response to the rule attribute being satisfied, the second determining module is further configured to determine a next node relative to the fault propagation node according to the selected fault propagation graph, determine a node type of the next node, and determine the suspected fault root cause node according to the node type of the next node;
in response to the rule attribute being not satisfied, the second determining module is further configured to stop determining the next node relative to the fault propagation node;
in response to the node type of the node being the fault propagation node and the verifiability attribute of the node being “unverifiable”, the second determining module is further configured to determine the next node relative to the fault propagation node according to the selected fault propagation graph, determine the node type of the next node, and determine the suspected fault root cause node according to the node type of the next node.

### Claim 13 (depends on 8)

The device for determining a root cause of fault of claim 8, further comprising a fault propagation graph generation module configured to
acquire, from a preset database of knowledge and experience of operation and maintenance, fault information comprising fault types, fault root cause types, events related to the fault types, and association relationships among the fault types, the fault root cause types and the events related to the fault types, and
generate a fault propagation graph for each of the fault types according to the fault information, wherein the fault propagation graph is a knowledge graph of fault propagation.

### Claim 14 (depends on 9)

The device for determining a root cause of fault of claim 9, further comprising a Bayesian model building module configured to
delete a node whose verifiability attribute is “unverifiable” in the fault propagation graph,
delete association relationships related to the deleted node,
establish, according to the deleted association relationships, an association relationship between nodes associated with the deleted node, so as to obtain a preprocessed fault propagation graph, and
respectively convert all nodes and association relationships in the preprocessed fault propagation graph into nodes and association relationships in the Bayesian model, so as to build the Bayesian model.

### Claim 15 (depends on 1)

A server, comprising:
one or more processors; and
a storage device having one or more programs stored thereon,
wherein, when the one or more programs are executed by the one or more processors, the one or more processors perform the method for determining a root cause of fault of claim 1.

### Claim 16 (depends on 1)

A non-transitory computer-readable medium having a computer program stored thereon, wherein, when the computer program is executed by a processor, the processor performs the method for determining a root cause of fault of claim 1.