Hierarchical Network Analytics for Distributed Topology Management

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Managing complex network topologies and their dynamic states in large distributed computing platforms is challenging due to the difficulty in understanding and analyzing network data effectively.

Innovation Solution

A hierarchical network analytics system that collects and analyzes data from network devices, performing first and second-level analyses based on filtered data sets to provide insights into network performance, topology, and issues, using a dashboard service that integrates data collection, analysis, and visualization modules.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Difficulty of detecting and measuring

If hierarchical analysis levels are implemented to manage complex network data, then network analysis capability is improved, but system complexity increases

Engineering Contradiction:
Improvenetwork analysis capabilityVSAvoidsystem complexity
Core Design Contradiction:
Difficulty of detecting and measuringVSDevice complexity

Solution Approach 1:

The patent divides network analysis into multiple hierarchical levels (first-level analyses for individual network devices and second-level analyses for network-wide patterns). This segmentation allows complex network data to be processed in manageable chunks at different levels, improving analysis capability while keeping each level's complexity manageable through structured organization of analysis tasks.

Inventive Principle:
Principle #1Segmentation

2Loss of information

If multiple levels of network analysis are performed, then insight depth is improved, but processing time increases

Engineering Contradiction:
Improveinsight depthVSAvoidprocessing time
Core Design Contradiction:
Loss of informationVSLoss of time

Solution Approach 1:

The patent performs first-level analyses on individual network devices before conducting second-level network-wide analyses. This preliminary action at the first level prepares and structures data in advance, so that when second-level analyses are performed, the processing is more efficient. The hierarchical structure ensures that foundational analysis is completed beforehand, reducing redundant processing time while maintaining deep insight generation.

Inventive Principle:
Principle #10Preliminary action

3Loss of information

If comprehensive network data collection is implemented, then data completeness is improved, but data processing load increases

Engineering Contradiction:
Improvedata completenessVSAvoiddata processing load
Core Design Contradiction:
Loss of informationVSProductivity

Solution Approach 1:

The patent segments network data collection and processing into first-level analyses for individual devices and second-level analyses for aggregate network patterns. This segmentation allows comprehensive data collection across the entire network while distributing the processing load across multiple hierarchical levels, preventing any single processing unit from being overwhelmed by the total data volume.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent performs analysis on subsets of network data at each hierarchical level rather than processing all data uniformly. First-level analyses process data for individual devices, and second-level analyses process aggregated results from multiple first-level analyses. This partial action approach maintains data completeness while significantly reducing the processing load at each level compared to processing all data in a single pass.

Inventive Principle:
Principle #16Partial or excessive action

Data Source

PatentUS11677635B2Hierarchical network analysis service
Publication Date: 2023.06.13 MICROSOFT TECHNOLOGY LICENSING LLC
  • US11677635B2 patent drawing
  • US11677635B2 patent drawing
  • US11677635B2 patent drawing

AI summary

A hierarchical network analytics system operated by a computing device or system is described. In some example techniques, the analytics system may determine results of a plurality of first level analyses each based at least in part on results of a respective plurality of data queries that return respective subsets of a plurality of types of network data. The analytics system may determine a result of a second level analysis based at least in part on results of the plurality of first level analyses.