Radio Access Network KPI Anomaly Detection After Software Updates
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current technologies fail to quickly and accurately detect network performance anomalies caused by software upgrades in radio access network devices, leading to potential degradation in network performance and inefficient use of wireless resources.
Innovation Solution
A computerized system utilizing a machine learning model to predict anomalies in radio access network devices by analyzing key performance indicators (KPIs) and initiating corrective actions such as restarting or rolling back software updates based on anomaly detection.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual review of alarms is performed to detect network performance degradation, then detection accuracy may be improved, but detection time increases significantly and may take days or weeks
Solution Approach 1:
The patent replaces manual alarm review (mechanical human operation) with an automated machine learning system that processes KPI data. The anomaly risk model automatically detects performance degradation by analyzing KPI trends and comparing them against baseline thresholds, eliminating the time-consuming manual review process while maintaining or improving detection accuracy through systematic algorithmic analysis.
Solution Approach 2:
The patent introduces KPI data as an intermediary between network operations and anomaly detection. Instead of directly reviewing alarms, the system monitors KPI metrics (such as throughput, latency, error rates) that serve as leading indicators of performance degradation. This intermediary approach enables earlier and more accurate detection by focusing on quantitative performance measures rather than reactive alarm analysis.
2Productivity
If software upgrades are implemented to improve network performance, then network capabilities are enhanced, but network performance degradation may occur as a side effect
Solution Approach 1:
The patent implements continuous feedback monitoring of KPI metrics after software upgrades are deployed. The anomaly risk model constantly compares current KPI performance against baseline thresholds and historical data, providing real-time feedback on whether the upgrade is causing performance degradation. This feedback mechanism enables operators to detect issues early and take corrective action before reliability is significantly impacted.
Solution Approach 2:
The patent performs preliminary analysis by establishing baseline KPI thresholds before software upgrades and continuously monitoring deviations from these baselines. The system proactively identifies anomalies that indicate potential reliability issues before they escalate into major network failures, enabling preventive maintenance and rollback decisions before significant service degradation occurs.
3Productivity
If automated anomaly detection is implemented using machine learning models, then detection speed and accuracy are improved, but system complexity increases
Solution Approach 1:
The patent employs a universal anomaly risk model that can detect multiple types of network anomalies across different radio access network devices using the same machine learning framework. The system processes various KPI metrics (throughput, latency, signal quality, error rates) through a single unified model, reducing the need for multiple specialized detection systems and thereby managing complexity while maintaining high detection capabilities.
Data Source
AI summary
Computerized systems and methods are provided that utilize a machine learning model to predict anomalies for a radio access network site device and taking corrective action. After a radio access network site device has received a software update, key performance indicators are extracted from the data from the radio access network site device. The key performance indicators are utilized with an anomaly risk model to determine if one or more of the key performance indicators exceeds a baseline threshold for the one or more radio access network devices. Responsive to the one or more key performance indicators exceeding a baseline threshold, a corrective action is initiated modifying the one or more radio access network devices.