UE L1-RSRP Reporting for AI Beam Prediction Reliability
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional beam management in 5G NR lacks sufficient data for reliable training and inference of AI/ML models due to incomplete reporting of L1-RSRP or L1-SINR measurements by UEs, inhibiting effective beam prediction at base stations.
Innovation Solution
UEs enhance reporting by providing detailed signaling to base stations about their L1-RSRP or L1-SINR calculation schemes, including whether instantaneous, the specific time instances, the application of filtering, and the time window, enabling more reliable data collection for AI/ML-based beam prediction.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If UEs provide detailed signaling about L1-RSRP/L1-SINR calculation schemes, then the accuracy and reliability of beam prediction is improved, but the reporting overhead and complexity increase
Solution Approach 1:
The patent introduces new parameters to the measurement report including measurement behavior indicators, time instances, filtering information, and time windows. These additional parameters enable the base station to accurately reconstruct the UE's measurement process and improve beam prediction reliability while maintaining manageable reporting complexity through structured parameter definitions
2Measurement precision
If UEs report comprehensive measurement data including time instances and filtering details, then the quality of AI/ML model training data is improved, but the data collection overhead increases
Solution Approach 1:
The patent performs preliminary organization of measurement data by structuring reports with predefined fields for time instances, filtering information, and time windows. This preliminary structuring enables efficient data collection and processing, improving measurement data quality for AI/ML training while minimizing overhead through pre-defined data formats
3Manufacturing precision
If the base station collects detailed measurement behavior information from UEs, then the accuracy of beam determination is improved, but the processing complexity at the base station increases
Solution Approach 1:
The patent segments the measurement behavior information into distinct components including measurement behavior indicators, time instances, filtering information, and time windows. This segmentation allows the base station to process each component systematically using dedicated algorithms, improving beam determination accuracy while managing processing complexity through modular data handling
Data Source
AI summary
Aspects are provided which allow a UE to provide detailed reporting to a base station regarding the UE's measurement behaviors or calculation schemes that the UE applied to determine its signal quality measurements (e.g., L1-RSRPs or L1-SINRs). For instance, the UE may transmit a message indicating a measurement behavior of the UE. After the UE receives a reference signal associated with a CMR, the UE transmits a report indicating a L1 signal quality metric associated with the CMR. The L1 signal quality metric is based on the measurement behavior indicated in the message. As a result, the base station may have a more reliable collection of data to apply to an AI/ML model for base station-based beam prediction, improving beam management performance or other AI/ML-based beam management.