Correlation Destruction Pattern Analysis for Cross-Device Failure Detection

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing system analysis methods struggle to determine failure causes when there are changes in the correlation model or when devices of the same type but different are involved, as they require identical correlation models and failure locations for accurate determination.

Innovation Solution

A system analysis device that generates aggregated destruction patterns by grouping correlation destruction patterns of the same type and calculates similarity between these patterns and new detection results, allowing for versatile state detection even with changes in the correlation model or different devices.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If correlation destruction patterns are compared to determine failure causes, then failure cause determination accuracy is improved, but the method becomes inapplicable when there are changes in the correlation model or when different devices of the same type are involved

Engineering Contradiction:
Improvefailure cause determination accuracyVSAvoidapplicability to different devices and model changes
Core Design Contradiction:
Measurement precisionVSAdaptability or versatility

Solution Approach 1:

The patent creates a reference correlation destruction pattern from historical failure data and uses it as a template to compare against new failure patterns. Instead of requiring exact matches between specific devices, the system copies the pattern-matching approach to identify similar failure modes across different devices and correlation models, enabling versatile failure cause determination while maintaining accuracy through pattern similarity comparison.

Inventive Principle:
Principle #26Copying

2Measurement precision

If the system requires identical correlation models for failure analysis, then determination accuracy is improved, but the system cannot handle distributed processing scenarios or model changes

Engineering Contradiction:
Improvefailure cause determination accuracyVSAvoidsystem configuration rigidity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent develops a universal failure analysis method that works across different correlation models and device configurations. By extracting failure patterns from historical data and comparing them with new failures using similarity metrics, the system achieves multi-functionality that handles both identical and changed correlation models, as well as different devices of the same type, without requiring exact model matching.

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

3Measurement precision

If specific failure locations must match for pattern comparison, then determination precision is improved, but the method fails when different devices experience similar failures

Engineering Contradiction:
Improvefailure cause determination precisionVSAvoidcross-device failure detection capability
Core Design Contradiction:
Measurement precisionVSAdaptability or versatility

Solution Approach 1:

The patent transforms the comparison approach by changing from exact location matching to pattern similarity comparison. Instead of requiring identical failure locations, the system uses parameters such as correlation destruction patterns, metric relationships, and failure characteristics to identify similar failures across different devices. This parameter transformation enables both precision through pattern matching and versatility across different device instances.

Inventive Principle:
Principle #35Parameter changes

Data Source

PatentEP2958023B1System analysis device and system analysis method
Publication Date: 2022.04.27 NEC CORP
  • EP2958023B1 patent drawingFigure 1
  • EP2958023B1 patent drawingFigure 2
  • EP2958023B1 patent drawingFigure 3

AI summary

In state detection of a system using a correlation destruction pattern, the versatility of the correlation destruction pattern is improved. A system analysis device (100) includes a correlation destruction pattern storage unit (113), an aggregated destruction pattern generation unit (104), and a similarity calculation unit (105). The correlation destruction pattern storage unit (113) stores a plurality of correlation destruction patterns (123) each of which is a set of correlations in which correlation destruction has been detected among correlations of pairs of metrics in a system. The aggregated destruction pattern generation unit (104) generates an aggregated destruction pattern (124) which is obtained by aggregating correlation destruction patterns (123) of the same type among the plurality of correlation destruction patterns (123). The similarity calculation unit (105) calculates and outputs a similarity between the aggregated destruction pattern (124) and a newly-detected correlation destruction pattern (123).