Radio Link Quality Prediction via Motion Data and Coverage Maps
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current cellular network technologies lack effective methods for predicting future radio link quality, leading to sub-optimal user experiences due to unpredictable connectivity issues, especially in environments with varying signal strength and mobility, which affects service level agreements and resource allocation.
Innovation Solution
A Link Quality Prediction Protocol (LQPP) that combines channel quality indicator reports with coverage maps, environmental data, and machine learning to predict radio link quality based on client motion and environmental conditions, enabling endpoints to adjust behavior proactively.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If cellular network technologies use current radio link monitoring methods, then network operation is maintained, but future radio link quality cannot be predicted leading to sub-optimal user experience
Solution Approach 1:
The system performs preliminary actions by collecting and storing historical radio link quality data, device location data, and network data before the actual link quality degradation occurs. This historical data is then used to train machine learning models that can predict future link quality, allowing endpoints to take proactive actions (such as buffering data or adjusting transmission parameters) before degradation happens, thereby resolving the contradiction between maintaining current network operation and predicting future quality.
2Reliability
If endpoints react to current link quality degradation, then connectivity issues are addressed, but service level agreements are violated due to reactive rather than proactive management
Solution Approach 1:
The system enables preliminary action by predicting future link quality degradation before it actually occurs. The machine learning model analyzes historical data and current conditions to forecast when link quality will deteriorate, allowing endpoints to proactively buffer data or adjust transmission parameters in advance. This eliminates the reactive delay inherent in traditional approaches, ensuring service level agreements are met by preparing before degradation happens rather than reacting after it occurs.
3Productivity
If network resources are allocated without future quality predictions, then current connectivity is maintained, but resource allocation is sub-optimal due to lack of predictive information
Solution Approach 1:
The system implements feedback mechanisms where actual link quality measurements are continuously compared with predicted quality values. This feedback loop allows the machine learning models to be continuously refined and improved. The feedback information is also used to adjust resource allocation decisions in real-time, ensuring that network resources are allocated optimally based on both historical patterns and actual performance, thereby improving both productivity and reliability simultaneously.
Data Source
AI summary
System and techniques for radio link quality prediction are described herein. A mobile device may receive a device registration. Motion data for the mobile device may then be obtained. A predicted path for the mobile device may be derived from the motion data. A set of predicted radio metrics for the mobile device along the predicted path mat be produced via a dynamic coverage map. The set of predicted radio metrics may then be transmitted.