Network Anomaly Detection via Time-Series Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current network quality monitoring systems struggle to identify the root cause of degradation in mobile communication networks, making it difficult for administrators to prioritize and solve issues effectively due to limitations in analyzing time-series information such as temporality, persistence, and repeatability of problems.
Innovation Solution
An electronic device is configured to detect anomalies in network quality indicators, perform time-series analysis on representative causes, and propose solutions by determining a first time interval for anomaly detection, identifying anomaly samples, and conducting a time-series analysis on associated features to understand trends, seasonality, and residual components.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If only quality indicator at a point in time is analyzed, then the state of entity or cell can be known, but time-series information such as temporality, persistence or repeatability cannot be identified
Solution Approach 1:
The system performs preliminary actions by collecting and storing quality indicator data over extended time periods before anomalies occur. This historical data accumulation enables subsequent time-series analysis to identify patterns, temporality, persistence, and repeatability of network issues, transforming isolated point-in-time measurements into comprehensive temporal understanding.
Solution Approach 2:
The patent transitions from single-point time analysis to multi-dimensional time-series analysis by introducing temporal dimensions (when did it occur, how long did it last, frequency of recurrence). This dimensional expansion allows the system to analyze quality indicators not just at a snapshot in time, but across multiple time dimensions, revealing patterns and characteristics that are invisible in point-in-time data alone.
2Reliability
If comprehensive time-series analysis is performed, then root cause can be identified, but system complexity and processing time increase
Solution Approach 1:
The system segments the comprehensive time-series analysis into distinct analytical modules: anomaly detection module, pattern recognition module, causality analysis module, and solution generation module. Each module processes specific aspects of the data independently, then integrates findings. This segmentation reduces overall system complexity by allowing modular development, testing, and optimization of each component while maintaining comprehensive analysis capability.
Solution Approach 2:
The system implements feedback mechanisms where analysis results feed back into refined detection and analysis processes. By continuously learning from historical resolutions and updating detection models, the system improves root cause identification accuracy over time while adapting its complexity management strategies. The feedback loop allows the system to optimize its analytical depth based on actual problem resolution effectiveness.
3Loss of information
If detailed root cause analysis is performed, then comprehensive understanding is achieved, but troubleshooting efficiency decreases
Solution Approach 1:
The system performs self-service by automatically executing time-series analysis, identifying patterns, determining root causes, and generating troubleshooting recommendations without requiring manual intervention. The automated anomaly detection and causality analysis algorithms continuously monitor network data, eliminating the need for administrators to manually analyze extensive time-series data while still achieving comprehensive root cause understanding.
Solution Approach 2:
The system performs preliminary analytical actions by pre-processing and storing quality indicator data in optimized formats before actual troubleshooting is needed. Historical data is pre-organized with metadata about temporal patterns, enabling rapid retrieval and analysis during incidents. This preliminary preparation significantly reduces the time required for root cause analysis during actual troubleshooting operations.
Data Source
AI summary
A method of solving a problem of a network and an electronic device for performing the method are provided. The electronic device includes at least one processor. The processor may determine a representative cause representing a cause of each of anomaly samples of a quality indicator indicating a quality of the network. The processor may perform a time-series analysis on an indicator associated with the representative cause. The processor may propose a solution corresponding to the representative cause and a result of the time-series analysis.