On-Demand Fault Reduction Framework for Network Alarm Correlation

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Network Operation Centers face challenges in effectively monitoring and managing complex networks due to the flood of alarms generated by network devices, which can overwhelm operators and hinder timely identification of specific faults.

Innovation Solution

The On-Demand Fault Reduction Framework provides a graphical user interface for users to design correlation rules, allowing for aggregation, suppression, and root cause analysis of alarms, enabling the identification of faults and reducing unnecessary alarms through a correlation engine that processes and displays correlated alarms in a Fault Management dashboard.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Quantity of substance

If alarm aggregation and correlation are implemented, then the number of alarms is reduced, but the device complexity increases

Engineering Contradiction:
Improvenumber of alarmsVSAvoidcorrelation engine complexity
Core Design Contradiction:
Quantity of substanceVSDevice complexity

Solution Approach 1:

The correlation engine is divided into multiple independent components: event correlation module, alarm aggregation module, suppression module, and root cause analysis module. Each module handles specific correlation tasks independently, reducing the complexity of the overall system while maintaining comprehensive alarm reduction capabilities.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The correlation engine is designed as a universal platform that can handle multiple types of correlation operations (aggregation, suppression, root cause analysis) through a single integrated system. This multi-functional design reduces the need for separate specialized systems for each correlation type.

Inventive Principle:
Principle #6Universality (Multi-functionality)

2Measurement precision

If comprehensive alarm correlation is implemented, then fault identification accuracy is improved, but the processing time increases

Engineering Contradiction:
Improvefault identification accuracyVSAvoidalarm processing time
Core Design Contradiction:
Measurement precisionVSLoss of time

Solution Approach 1:

The system pre-configures correlation rules, aggregation criteria, and suppression patterns before alarm floods occur. Event correlation templates and root cause analysis models are prepared in advance, enabling the correlation engine to quickly match incoming alarms against pre-defined patterns rather than analyzing everything from scratch.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

An event correlation layer is introduced as an intermediary between raw alarm generation and fault identification. This layer aggregates similar events, filters out duplicates, and presents processed correlation results to the fault management system, reducing the time required for comprehensive analysis.

Inventive Principle:
Principle #24Intermediary (Mediator)

3Productivity

If real-time alarm correlation is implemented, then network monitoring effectiveness is improved, but the computational resources required increase

Engineering Contradiction:
Improvenetwork monitoring effectivenessVSAvoidcomputational resources
Core Design Contradiction:
ProductivityVSUse of energy by moving object

Solution Approach 1:

The correlation engine applies partial correlation processing to alarm streams by focusing on the most critical correlation operations first. Not all possible correlation rules are applied to every alarm - instead, the system prioritizes high-impact correlation tasks and applies less intensive processing to lower-priority alarms, reducing overall computational burden while maintaining monitoring effectiveness.

Inventive Principle:
Principle #16Partial or excessive action

Data Source

PatentEP3327637B1On-demand fault reduction framework
Publication Date: 2023.05.10 ACCENTURE GLOBAL SOLUTIONS LTD
  • EP3327637B1 patent drawingFigure 1
  • EP3327637B1 patent drawingFigure 2
  • EP3327637B1 patent drawingFigure 3

AI summary

A computer-implemented method for correlating alarms that are generated with relation to a computer network is disclosed. The alarms represent faults of the computer network and/or faults of components of the computer network. The computer-implemented method comprises: displaying, on a display device, a graphical user interface that allows a user to select a type of correlation between one or more alarms and a type of fault which causes the one or more alarms from among a plurality of types of correlation; receiving, via the graphical user interface, a user input indicating the type of correlation selected by the user; storing, in a database, at least one rule based on the selected correlation; receiving, by a server, a plurality of alarms generated with relation to the computer network; correlating, by a correlation engine, the plurality of alarms received by the server according to the at least one rule stored in the database to identify a fault which has caused the one or more alarms; and displaying, on the display device, the at least one identified fault.