Failure Effects Estimation Using Traffic Data and Logs
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional methods fail to efficiently estimate the effects of abnormalities on users in communication systems, as they cannot accurately determine the number of affected users, duration of impact, and changes in traffic, especially considering usage states and locations, which are crucial for prioritizing recovery and reducing operational expenses.
Innovation Solution
A failure effects estimating device that uses machine learning models to predict traffic changes and affected users by analyzing past logs and traffic data, enabling the estimation of failure effects such as the number of affected users, failure effects time, and network performance metrics like RTT and throughput.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional abnormality detection methods are used, then abnormalities can be detected, but the failure effects amount and user impact cannot be estimated
Solution Approach 1:
The system segments the analysis into two distinct modules: an abnormality detection module that identifies failures, and a failure effects estimation module that quantifies user impact. This segmentation allows each module to specialize in its function while working together to provide comprehensive failure analysis.
Solution Approach 2:
The system introduces traffic amount data as an intermediary element that bridges abnormality detection and failure effects estimation. By analyzing traffic patterns before and during abnormalities, the system can infer user impact without directly measuring it, solving the information loss problem.
2Measurement precision
If dependency relations among services and resources are defined in advance, then failure impact can be analyzed, but construction requires expertise and great amounts of time
Solution Approach 1:
The system enables automatic construction of dependency models by analyzing historical traffic data and abnormality patterns. Instead of requiring manual expert configuration, the system self-learns the relationships between services, sub-services, and resources from operational data, dramatically reducing construction time while maintaining accuracy.
Solution Approach 2:
The system performs preliminary analysis of traffic data and abnormality patterns during normal operations to pre-establish dependency relationships. When an abnormality occurs, the pre-analyzed data enables immediate failure effects estimation without requiring time-consuming model construction at the moment of failure.
3Measurement precision
If externally measuring services using simulated user terminals is performed, then service states and effect magnitude can be estimated, but the complexity and resource consumption increase
Solution Approach 1:
Instead of using complex simulated user terminals to measure service states, the system creates a virtual model of service dependencies and uses traffic amount data as a proxy measurement. This copying approach maintains measurement accuracy while avoiding the complexity of physical or virtual terminal simulations.
Solution Approach 2:
The system replaces the mechanical approach of using simulated user terminals with an information-based approach that analyzes existing traffic data and log information. This substitution eliminates the need for additional measurement infrastructure while achieving the same goal of estimating failure effects.
4Quantity of substance
If the number of terminals in base station range is predicted, then affected terminals can be estimated, but usage state differences between daytime and nighttime cannot be considered
Solution Approach 1:
The system dynamically adapts its analysis based on historical traffic patterns that capture usage state variations throughout the day. By learning from temporal patterns in the data, the system automatically adjusts its failure effects estimation to account for different usage states without requiring separate models for daytime and nighttime scenarios.
Solution Approach 2:
The system changes its analysis parameters based on the learned usage patterns from historical data. When an abnormality occurs, the system selects appropriate parameters and weighting factors based on the current time and historical usage patterns, enabling accurate estimation across different operational conditions without manual intervention.
Data Source
AI summary
A failure effects estimating device includes an input unit that inputs a log and a traffic amount obtained from a communication system when an abnormality occurs, an estimating unit that estimates a failure effects amount in the communication system, on the basis of the log and the traffic amount, and an output unit that outputs the failure effects amount estimated by the estimating unit.