Cellular Network Cell Anomaly Detection via KPI Change Points
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current automated monitoring arrangements in cellular communications networks fail to detect certain malfunctions, necessitating a more advanced approach to analyze cell operation and identify anomalies.
Innovation Solution
A computer-implemented method for analyzing cell performance by identifying change points in performance indicators, comparing them to reference cells, and determining relevance using statistical methods like binary segmentation and Z-score, to trigger corrective actions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If current automated monitoring arrangements are used, then basic network operation monitoring is achieved, but malfunctions are not detected and monitoring precision is insufficient
Solution Approach 1:
The monitoring approach segments performance data into individual performance indicators (KPIs) and identifies change points within each indicator's time series. By dividing the monitoring task into detecting change points in each KPI separately, the system achieves more precise detection of anomalies that would be missed in aggregate monitoring.
Solution Approach 2:
The system performs preliminary analysis by identifying change points and calculating their magnitudes before determining relevance. This preliminary detection of statistical deviations allows the system to prepare and prioritize anomalies for further investigation, improving both precision and reliability of malfunction detection.
2Loss of information
If multiple performance indicators are monitored continuously, then comprehensive network analysis is achieved, but data processing complexity increases
Solution Approach 1:
The system extracts only the critical information needed for anomaly detection - specifically, change points and their magnitudes from the continuous performance indicator data. By taking out only these key features rather than processing all raw data, the system maintains information completeness while reducing processing complexity.
Solution Approach 2:
The system applies partial action by focusing analysis on change points that exceed certain thresholds or represent significant deviations, rather than equally analyzing all data points. This selective approach to processing reduces complexity while maintaining comprehensive monitoring of important anomalies.
3Measurement precision
If change points are identified using statistical methods, then detection accuracy improves, but computational requirements increase
Solution Approach 1:
The system changes parameters by transforming continuous performance indicator data into discrete change point events with associated magnitudes. This parameter transformation allows statistical methods to be applied efficiently, improving detection accuracy while reducing computational energy requirements compared to continuous analysis.
Data Source
AI summary
A computer implemented method for analyzing operation of cells of a communications network is provided. The method includes: obtaining data comprising performance indicators data from a first cell and a group of reference cells for a selected time period;—identifying, for the first cell, first change point of a first performance indicator, calculating magnitude of the change of the first performance indicators at the first change point, and defining the calculated magnitude as a first magnitude; identifying, for the group of reference cells, a group of reference change points of the first performance indicator, calculating magnitudes of the changes of the group of reference change points, and defining the calculated magnitudes as a group of reference magnitudes; comparing the first magnitude to the group of reference magnitudes to determine the relevance of the change point of the first cell; and providing output indicating the first change point of the first cell in response to detecting that said change point is determined relevant, or otherwise, providing output indicating that no relevant change points are identified.