Network Counter Selection via Delta-Averages and Cross-Entropy
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In computer networking, identifying the most relevant features contributing to a state change in network components is challenging due to the large number of operational counters, making it difficult for network engineers to determine which counters are related to the state change, and assessing the quality of counter selections is not feasible.
Innovation Solution
The use of delta-averages to identify the most descriptive counters for a state change, combined with a cross-entropy based metric to quantify intelligibility and an optimization score, allows for the automatic selection of the most relevant counters, enabling network engineers to understand and address state changes effectively.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If network engineers manually select operational counters for monitoring, then they can focus on specific counters of interest, but they cannot monitor the full set of counters and thus miss counters that react to state changes
Solution Approach 1:
The system performs automatic counter selection and analysis without requiring manual intervention. The automated system identifies and ranks counters based on their relationship to state changes, eliminating the need for engineers to manually select counters while still focusing on the most relevant ones.
Solution Approach 2:
The system changes the parameter of counter selection from manual engineer judgment to automated data-driven ranking based on delta-averages and cross-entropy metrics. This transforms the selection criterion from subjective expertise to objective quantitative measurement.
2Loss of information
If all operational counters are monitored, then complete data is available for analysis, but the large volume of counters makes it difficult to determine which counters are related to state changes
Solution Approach 1:
The system extracts and isolates the most relevant counters from the complete set of monitored counters by calculating delta-averages and applying cross-entropy-based ranking. This separates the signal (relevant counters) from the noise (irrelevant counters) while maintaining comprehensive monitoring.
Solution Approach 2:
The system segments the large set of counters into ranked groups based on their relevance to state changes. By dividing the counters into ordered categories, the system makes the analysis manageable while preserving information from all counters.
3Reliability
If network engineers rely on domain expertise to interpret operational data, then they can make informed decisions, but this approach is time-consuming and requires extensive manual intervention
Solution Approach 1:
The system replaces the mechanical process of manual expert analysis with an automated computational system that uses delta-averages and cross-entropy metrics to identify relevant counters and their relationships to state changes, significantly reducing troubleshooting time while maintaining reliability.
Solution Approach 2:
The system introduces an intermediary automated analysis layer between the raw operational data and the network engineer. This intermediary performs the time-consuming analysis work of identifying relevant counters and their relationships, leaving the engineer to focus on decision-making based on pre-processed insights.
4Ease of manufacture
If traditional monitoring approaches are used, then simple protocols like SNMP can retrieve operational data, but there is no way to assess the quality of counter selections
Solution Approach 1:
The system introduces feedback through cross-entropy-based quality assessment metrics that evaluate how well selected counters represent state changes. This feedback mechanism allows the system to measure and optimize counter selection quality, transforming the process from subjective choice to measurable performance.
Data Source
AI summary
Techniques and mechanisms for automatically identifying counters/features of a network component that are related to a state change (or event) for the network component or for the network itself. For example, using data obtained from the network component around a time of the state change, delta-averages for the counters/features around the time of the state change may be determined. The delta-averages may be utilized to determine which counters/features are most descriptive for a particular state change. Determining which counters/features are most descriptive may also include determining which counters/features are most relevant, i.e., counters/features that contribute most to preserving the manifold structure of the original data or counters/features with the highest or lowest correlation with the other counters/features in the data set. Thus, the techniques described herein provide for an approach to distill which counters/features contribute the most to a particular state change from a data driven perspective.