Beam Pairing Prediction Using Spatial Reference Beam Information
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The challenge of identifying suitable beam pairs between a gNB and a UE for data transmission in high-frequency wireless networks is exacerbated by large numbers of potential beam pairs, leading to significant network signaling overhead, increased latency, and inefficient use of radio resources due to limited information about Rx and Tx beam configurations.
Innovation Solution
A method where a first radio node provides beam-related information, including spatial properties of reference signal beams, to a second node for predicting beam pair quality, reducing the need for extensive measurements and reporting.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If beam management procedures are performed to establish and maintain suitable beam pairs, then reliable data transmission is achieved, but network signaling overhead and latency increase significantly
Solution Approach 1:
The patent applies preliminary action by performing beam pairing prediction using machine learning models before actual data transmission. The system predicts suitable beam pairs in advance based on historical measurement data and spatial correlation information, so that when data transmission is needed, the predicted beam pairs can be immediately activated without performing full beam management procedures, thus reducing latency while maintaining reliability
Solution Approach 2:
The patent segments the beam management process into two parts: (1) a prediction phase using machine learning to identify candidate beam pairs, and (2) a verification phase using actual measurements. This segmentation allows the system to use the ML model to filter out unlikely beam pairs, reducing the number of measurements needed and thereby reducing overall latency while maintaining reliable beam pair selection
2Measurement precision
If extensive beam measurements and reporting are performed to identify suitable beam pairs, then accurate beam pairing is achieved, but network signaling overhead increases
Solution Approach 1:
The patent applies partial action by using the machine learning model to predict beam pair qualities for all potential beam pairs, then selecting only the top predicted candidates for actual measurement and reporting. This means measurements are performed on a subset (partial action) rather than all beam pairs, reducing signaling overhead while maintaining accurate beam pairing through the combination of ML prediction and targeted measurements
Solution Approach 2:
The machine learning model acts as an intermediary between the raw measurement data and the final beam pair selection. It processes spatial correlation information and historical data to predict beam pair qualities, reducing the need for extensive direct measurements and reporting. The ML model mediates the information flow, allowing accurate beam pairing with reduced signaling overhead
3Device complexity
If limited information about Rx and Tx beam configurations is used, then processing complexity is reduced, but beam pairing prediction accuracy deteriorates
Solution Approach 1:
The patent applies universality by designing a machine learning model that can process multiple types of input information (spatial correlation data, historical measurement results, beam configuration parameters) through a unified framework. The same ML model structure handles different input combinations, reducing processing complexity while maintaining prediction accuracy through the model's ability to learn from diverse data sources
Data Source
AI summary
A method (1900) by a first radio node (205, 210) includes receiving (1902), from a second radio node (205, 210), beamforming information comprising at least one parameter related to at least one spatial property of at least one reference signal beam. In particular, the beamforming information related to the at least one spatial property may include at least one of a beam configuration, a spatial correlation between two or more reference signal beams, and/or a capability related to beamforming at the second radio node.