Network Control via Association Rule Change Pattern Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current network management is largely reactive, relying on critical alerts to identify and address issues, which can lead to major outages, and often fails to detect underlying problems in other network areas that may cause issues, requiring dedicated teams to monitor hundreds or thousands of nodes.
Innovation Solution
A method that analyzes network operating data to identify change patterns such as trends, spikes, and stabilities in association rule confidence and support over time, allowing for proactive reconfiguration of the network by matching these patterns against predetermined ones and using interestingness parameters to filter and rank change patterns for human analysis.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If network monitoring relies on critical alerts from individual nodes, then the system can detect obvious failures, but it cannot identify underlying problems in other network areas that may cause issues
Solution Approach 1:
The patent segments network monitoring into multiple analytical layers: individual node metrics, association rules between node pairs, and change patterns in rule confidence/support values. This segmentation allows detection of subtle changes in relationships between nodes before they manifest as critical failures, addressing the limitation of traditional single-node monitoring.
Solution Approach 2:
The system performs preliminary analysis by continuously tracking changes in association rule confidence and support values over time periods. This preliminary detection identifies deteriorating relationships between nodes before they cause critical failures, enabling proactive intervention rather than reactive response to alerts.
2Reliability
If a large network has hundreds or thousands of nodes monitored individually, then comprehensive coverage is achieved, but a dedicated team of specialists is required
Solution Approach 1:
The patent merges individual node monitoring data into association rules that capture relationships between nodes. By combining metrics from multiple nodes into unified association rules and then analyzing change patterns in rule confidence and support, the system reduces the complexity of monitoring thousands of individual nodes while maintaining comprehensive coverage.
Solution Approach 2:
The association rule framework serves multiple functions simultaneously: it monitors individual node performance, detects relationships between nodes, identifies change patterns, and prioritizes alerts. This multi-functionality reduces the need for specialized monitoring systems for different network aspects.
3Productivity
If network management is reactive based on critical alerts, then response to obvious failures is enabled, but problems are only identified once they become critical causing major outages
Solution Approach 1:
The system performs preliminary detection by continuously analyzing change patterns in association rule confidence and support values. This enables identification of deteriorating network conditions before they reach critical thresholds, allowing proactive reconfiguration or intervention to prevent outages rather than reacting after failures occur.
Solution Approach 2:
The system establishes feedback loops by continuously monitoring change patterns in association rules and using this information to trigger alerts or automated responses. The feedback mechanism transforms static network configuration into a dynamic system that adapts to changing conditions, enabling timely responses to emerging problems.
Data Source
AI summary
Patterns of change in system operating parameters are identified which may be used to identify unexpected operational conditions and to trigger an appropriate alert or action. A network is controlled using operating data for the network. Network operating data is received and divided into sequential time periods. A structural description such as an association rule is determined for the received data in each time period and a change pattern in the determined structural description is identified over the time periods. The network is then controlled using a predetermined action corresponding to a predetermined structural description change pattern in response to the identified structural description change pattern matching the predetermined structural description change pattern.