Automated Alarm Correlation Data Generation Through Time-Window Grouping

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing methods for generating correct data for alarm correlation in network monitoring are burdensome for maintenance personnel, requiring manual rule definition and large amounts of learning data, and are not fully automated.

Innovation Solution

A device and method that automatically group alarms with close occurrence and recovery times, using a correlation unit to associate alarms within a time width and generate correct data with event identification for each alarm, reducing the need for manual intervention.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If manual correlation of alarms is performed by maintenance personnel, then correlation accuracy can be maintained, but the workload and time consumption increase significantly

Engineering Contradiction:

Improvecorrelation accuracyVSAvoidtime consumption

Core Design Contradiction:

Measurement precisionVSLoss of time

Solution Approach 1:

The system enables self-service automation where the alarm correlation process is performed automatically by the system itself using machine learning models, eliminating the need for manual intervention while maintaining high correlation accuracy through automated pattern recognition and event grouping

Inventive Principle:

Principle #25Self-service

Solution Approach 2:

The manual mechanical process of correlating alarms by maintenance personnel is replaced with an automated electronic system using machine learning algorithms that can process and correlate alarms rapidly without human intervention, significantly reducing time consumption while maintaining accuracy

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

2Productivity

If machine learning is used for alarm correlation automation, then productivity increases, but large amounts of correct training data are required

Engineering Contradiction:

Improveautomation efficiencyVSAvoidtraining data volume

Core Design Contradiction:

ProductivityVSQuantity of substance

Solution Approach 1:

The system performs preliminary actions by automatically generating synthetic correct alarm data that simulates real alarm correlations, preparing training data in advance without requiring manual collection of large volumes of real-world correlated alarm data, thus enabling machine learning model training with sufficient productivity

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The system creates copies of alarm data by generating synthetic alarm correlation data that mimics real alarm patterns and relationships, providing sufficient training data for machine learning without requiring actual collection of large volumes of real correlated alarm data from production environments

Inventive Principle:

Principle #26Copying

3Extent of automation

If rule-based correlation methods are used, then automation can be partially achieved, but rule definition requires manual intervention

Engineering Contradiction:

Improveautomation levelVSAvoidrule definition complexity

Core Design Contradiction:

Extent of automationVSEase of operation

Solution Approach 1:

The system changes the approach from static rule definition to dynamic parameter-based correlation by using machine learning models that automatically learn correlation parameters from data, achieving full automation without requiring manual rule definition while maintaining ease of operation through automated parameter optimization

Inventive Principle:

Principle #35Parameter changes

Data Source

PatentUS12388706B2Correct answer data generation apparatus, correct answer data generation method and correct answer data generation program

Publication Date: 2025.08.12 KPS GLOBAL LLC

AI summary

A correct data generation device 1 includes: an acquisition unit 11 that acquires alarm information output from a plurality of devices; a correlation unit 13 that associates, from the alarm information, alarms whose occurrence times are within a first time width and recovery times are within a second time width, as a group of alarms that have occurred by the same event; and a generation unit 14 that generates correct data in which identification information of the event is set for each alarm of the group of alarms that has been associated.