Context-Based Anomaly Detection for Cellular Network KPIs
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The complexity and size of modern cellular networks make it challenging for operators to effectively monitor and manage network performance, with a large volume of metadata generated by network devices, requiring continuous monitoring of key performance indicators (KPIs) and quality indicators (KQIs) across thousands of base stations, which is daunting for human engineers.
Innovation Solution
A computing device in a communication system performs correlation analysis to assign contexts to time intervals of data, groups historic time-series data based on these contexts, computes anomaly scores by comparing new data to historic data, and indicates anomalies based on thresholds, enabling context-based multivariate anomaly detection and aggregate anomaly scoring.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If human engineers manually monitor network performance benchmarks across thousands of base stations, then service quality can be maintained, but the monitoring task becomes daunting and inefficient
Solution Approach 1:
The system enables self-service by implementing automated anomaly detection that operates independently without human intervention. The computing device automatically assigns contexts to time intervals, groups historic data, computes anomaly scores, and indicates anomalies by comparing new data against historical patterns, allowing the network monitoring system to serve itself
Solution Approach 2:
The patent replaces the mechanical system of manual human monitoring with an automated computational system. The computing device uses algorithmic processes including correlation analysis, data grouping, anomaly score computation, and threshold-based indication to substitute human engineers in the monitoring task, dramatically improving ease of operation while maintaining reliability
2Productivity
If the network size and complexity increase to accommodate more base stations and devices, then network coverage and capacity improve, but management becomes highly challenging and costly
Solution Approach 1:
The system applies segmentation by dividing the complex network management task into distinct processing stages: assigning contexts to time intervals, grouping historic time-series data by context, computing anomaly scores for each context, and indicating anomalies based on threshold comparisons. This segmented approach makes managing large-scale networks more tractable
Solution Approach 2:
The computing device performs multiple functions within a single integrated system: it conducts correlation analysis, assigns contexts, groups data, computes anomaly scores, and indicates anomalies. This multi-functional approach consolidates various management tasks into one universal system, reducing overall management complexity and cost
3Loss of information
If a large volume of metadata is generated and continuously monitored to maintain service quality, then network performance can be tracked, but the burden on human engineers increases
Solution Approach 1:
The system performs self-service by automatically processing the large volume of metadata without human intervention. The computing device continuously assigns contexts, groups data, computes anomaly scores, and indicates anomalies, eliminating the need for human engineers to manually analyze the metadata while maintaining complete network performance tracking
Solution Approach 2:
The system ensures continuous monitoring by continuously performing all processing steps: assigning contexts to incoming time intervals, grouping historic data, computing anomaly scores, and indicating anomalies. This continuous automated action maintains complete information tracking while eliminating time burden on human engineers
Data Source
AI summary
Methods and apparatuses for automating configuration management in cellular networks. A method of a computing device comprises: assigning, based on a correlation analysis, contexts to different time intervals of data, wherein the correlation analysis is performed based on historic time-series data; grouping, based on the assigned contexts, the historic time-series data; identifying context and compute an anomaly score comparing new data and the grouped historic-time series data of the context; indicating an event of anomaly based on a determination that the computed anomaly score exceeds a first threshold that is identified based on a function of per-context data; and computing, based on the event of the anomaly, an aggregate anomaly score or indicate using a value of mean or moving average of a set of latest anomaly scores, for a context-based multivariate anomaly detection.