Telemetry Data Subdivision for Network Anomaly Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
High-dimensional telemetry data from computer networks is complex and impractical to manually analyze, making it difficult to uncover specific patterns that could indicate network issues or outages, as most network problems have underlying patterns associated with common root causes.
Innovation Solution
A network assurance service subdivides telemetry data into subsets by device type, time period, and metric type, computes distribution percentiles, identifies outlier subsets by comparing these summaries, and reports insights to a user interface, while adjusting subsets based on user feedback to enhance pattern discovery and anomaly detection.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual analysis of telemetry data is performed, then analysis accuracy can be maintained, but analysis efficiency deteriorates due to high dimensionality and complexity
Solution Approach 1:
The patent segments high-dimensional telemetry data into multiple lower-dimensional subsets using clustering algorithms. Each subset contains data points with similar characteristics, making them more manageable for analysis. This segmentation maintains analytical precision while improving efficiency by breaking down the complex high-dimensional problem into simpler components that can be processed more quickly.
2Reliability
If traditional analytics are applied to captured network information, then network health assessment is achieved, but the ability to discover hidden patterns deteriorates due to data volume and dimensionality
Solution Approach 1:
The patent extracts meaningful patterns from high-dimensional telemetry data by identifying and removing redundant dimensions. The system extracts key features that represent network health while discarding unnecessary data that contributes to dimensionality but not to pattern discovery. This extraction process preserves the ability to assess network health while recovering lost pattern discovery capability.
Solution Approach 2:
The patent transforms high-dimensional telemetry data into lower-dimensional representations through clustering and dimensionality reduction techniques. By projecting data from high-dimensional space into lower-dimensional clusters, the system maintains the essential information needed for pattern discovery while reducing the overwhelming complexity that causes information loss.
3Adaptability or versatility
If telemetry data is collected from multiple sources and dimensions, then network monitoring comprehensiveness is improved, but data complexity and analysis difficulty worsen
Solution Approach 1:
The patent merges multiple telemetry data sources and dimensions into unified cluster representations. Instead of analyzing each data source separately, the system combines them into integrated clusters that capture the essential relationships across all dimensions. This merging reduces analysis complexity while preserving the comprehensive monitoring capability by maintaining the interconnectedness of multiple data sources within each cluster.
Data Source
AI summary
In one embodiment, a network assurance service that monitors a plurality of networks subdivides telemetry data regarding devices located in the networks into subsets, wherein each subset is associated with a device type, time period, metric type, and network. The service summarizes each subset by computing distribution percentiles of metric values in the subset. The service identifies an outlier subset by comparing distribution percentiles that summarize the subsets. The service reports insight data regarding the outlier subset to a user interface. The service adjusts the subsets based in part on feedback regarding the insight data from the user interface.