Self-Healing Diagnosis Knowledge Sharing for Network Anomalies
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The automated diagnosis of anomalies in self-organizing networks is challenging due to their distributed and heterogeneous nature, making it difficult to collect statistically meaningful data sets for rare fault states, which hinders reliable root-cause analysis and knowledgebase maintenance.
Innovation Solution
The implementation of a self-healing diagnosis system using transfer learning and case-based reasoning, where local diagnosis systems share knowledge with a central diagnosis system to update and refine cluster models, enabling dynamic diagnosis and corrective actions based on previous anomaly patterns.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Extent of automation
If automated diagnosis is implemented in distributed networks, then diagnostic capability is improved, but data collection reliability deteriorates due to rare fault states
Solution Approach 1:
The patent combines data from multiple local diagnosis systems into a centralized knowledge base. By merging datasets across different network locations and fault types, the system accumulates sufficient statistical samples for rare fault states that would be impossible to observe at any single location, thereby resolving the reliability issue while maintaining automated diagnosis.
Solution Approach 2:
The centralized knowledge base serves multiple local diagnosis systems simultaneously, providing a universal repository that benefits all participants. Each local system contributes data and receives improved diagnostic models, creating a multi-functional system that addresses both local and global diagnostic needs.
2Speed
If knowledgebase is maintained locally at each diagnosis system, then response speed is improved, but knowledge completeness deteriorates
Solution Approach 1:
The system segments the knowledge base into local and centralized components. Local diagnosis systems maintain lightweight models for immediate response, while the centralized system stores comprehensive knowledge. This segmentation allows fast local decisions while preserving complete information centrally.
Solution Approach 2:
The centralized knowledge base pre-processes and consolidates diagnostic knowledge from all sources before distributing updated models to local systems. This preliminary action ensures that local systems receive ready-to-use, comprehensive knowledge without needing to process raw data themselves, maintaining both speed and completeness.
3Measurement precision
If centralized knowledge base is used, then diagnostic accuracy is improved, but system complexity increases
Solution Approach 1:
The patent introduces a centralized knowledge base as an intermediary between local diagnosis systems and the comprehensive diagnostic data. This intermediary consolidates information from multiple sources and provides processed knowledge to local systems, improving accuracy while shielding local systems from the complexity of managing distributed data collection and aggregation.
4Reliability
If more data is collected from multiple sources, then statistical significance is improved, but data management complexity increases
Solution Approach 1:
The centralized knowledge base merges data from multiple local diagnosis systems into a unified repository. By combining datasets centrally, the system achieves statistical significance for rare events without requiring each local system to manage complex distributed data infrastructure, thus reducing overall data management complexity while improving reliability.
Data Source
AI summary
According to an aspect, there is provided a local diagnosis system comprising means for performing the following. The local diagnosis system detects one or more anomaly events associated with a communications network. Each anomaly event defines an anomaly pattern describing a data point in a performance indicator space. Then, the local diagnosis system updates one or more local cluster models to incorporate the one or more anomaly patterns within complexity constraints. Each of the one or more local cluster models corresponds to a different diagnosis label defining a diagnosis. In response to failing according to one or more per-defined criteria to incorporate, in the updating, the one or more anomaly patterns to the one or more local cluster models, the local diagnosis system forwards at least the one or more local cluster models and one or more associated diagnosis labels to a central diagnosis system for further diagnosis.