OpenRAN Anomaly Propagation Prediction Across Correlated Cells
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In modern cellular networks with disaggregated components managed by multiple vendors, detecting and understanding the propagation of anomalies is difficult, leading to reactive responses that cause delays and excessive troubleshooting, affecting network performance and user satisfaction.
Innovation Solution
A system for predicting anomaly propagation using spatio-temporal correlation across cells, incorporating cross correlation and external data sources like user movement and weather data to proactively allocate resources and mitigate anomalies.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If reactive anomaly metric analysis is used, then response to anomalies can be made, but excessive delay in addressing anomalies occurs
Solution Approach 1:
The system performs preliminary actions by proactively predicting anomaly propagation paths and affected cells before anomalies actually occur. Using spatio-temporal correlation analysis of historical anomaly data, the system identifies likely propagation routes and prepares mitigation strategies in advance, enabling rapid response when anomalies occur rather than reacting with delay after detection
Solution Approach 2:
The system implements feedback mechanisms by continuously monitoring current anomaly metrics, comparing them against predicted propagation patterns, and adjusting resource allocation dynamically. This closed-loop feedback enables the system to respond proactively to emerging anomalies by detecting early signs that match predicted propagation patterns and initiating mitigation before full propagation occurs
2Loss of information
If comprehensive anomaly propagation analysis is performed across the network, then anomaly understanding improves, but system complexity increases
Solution Approach 1:
The system segments the network into propagation zones based on spatio-temporal correlation analysis, dividing the complex network into manageable regions with similar anomaly propagation characteristics. This segmentation allows focused analysis on relevant zones rather than treating the entire network uniformly, reducing computational complexity while maintaining comprehensive propagation understanding
Solution Approach 2:
The system changes parameters by dynamically adjusting correlation thresholds, analysis time windows, and resource allocation levels based on current network conditions and anomaly patterns. This parameter adaptation enables the system to scale its analysis complexity according to actual needs, performing comprehensive analysis when necessary while reducing complexity during normal operation
Data Source
AI summary
Generally provided is a radio system that can comprise a lower-level controller, of a radio network, that is responsive to and provided at a level of hierarchy lower than an upper-level controller of the radio network, and the upper-level controller that generates a cell-level prediction of an anomaly at a first cell of a radio network based on cell-level data determined by the lower-level controller, wherein the upper-level controller generates a global prediction, based on the cell-level prediction, of propagation of the anomaly to a second cell of the radio network. The upper-level controller can generate the global prediction based on cross correlation of a first metric value, of the cell-level data, measured according to a defined metric, from the first cell and a second metric value, measured according to the defined metric, from the second cell of the radio network.