Spatial Filter Prediction for Low-Latency NR Beam Management
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The overhead and latency associated with traversing all combinations of transmit and receive beams in uplink and downlink beam management processes in millimeter wave frequency bands are significant, particularly in new radio (NR) systems, leading to inefficient beam management.
Innovation Solution
Implementing AI/ML models on terminal devices to predict uplink and downlink spatial filters based on measurement results, allowing for reduced overhead and latency in beam management by predicting optimal beam pairs.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If all combinations of transmit beams and receive beams are traversed to select optimal beams, then beam selection accuracy is improved, but overhead and latency are increased
Solution Approach 1:
The patent applies preliminary action by using downlink measurement results to predict uplink spatial filters before actual uplink transmission occurs. The terminal device performs downlink measurements and uses AI/ML models to predict the optimal uplink spatial filters in advance, so that when uplink transmission is needed, the prediction results can be directly used or with minimal adjustment, avoiding the need to traverse all beam combinations in real-time.
Solution Approach 2:
The patent introduces an intermediary mechanism by using downlink measurement results as an intermediate step to derive uplink spatial filter predictions. Instead of directly traversing uplink beams, the system uses downlink measurements as an intermediary that provides information about the channel state, which is then processed through AI/ML models to predict the optimal uplink spatial filters.
2Measurement precision
If all combinations of transmit beams and receive beams are traversed to select optimal beams, then beam selection accuracy is improved, but overhead and latency are increased
Solution Approach 1:
The patent applies preliminary action by using downlink measurement results to predict uplink spatial filters before actual uplink transmission occurs. The terminal device performs downlink measurements and uses AI/ML models to predict the optimal uplink spatial filters in advance, so that when uplink transmission is needed, the prediction results can be directly used or with minimal adjustment, avoiding the need to traverse all beam combinations in real-time.
Solution Approach 2:
The patent introduces an intermediary mechanism by using downlink measurement results as an intermediate step to derive uplink spatial filter predictions. Instead of directly traversing uplink beams, the system uses downlink measurements as an intermediary that provides information about the channel state, which is then processed through AI/ML models to predict the optimal uplink spatial filters.
3Loss of time
If AI/ML models are implemented on terminal devices for prediction, then overhead and latency are reduced, but device complexity is increased
Solution Approach 1:
The patent applies self-service by enabling the terminal device to perform beam prediction autonomously using locally deployed AI/ML models. The terminal device uses its own downlink measurement results as input to the model and generates predictions for uplink spatial filters without needing network assistance, thereby reducing latency and overhead while distributing the computational burden to the terminal.
Data Source
Figure 1~3
Figure 4~6
Figure 7~8
AI summary
Provided in the embodiments of the present application are wireless communication methods, and devices. A terminal device can support performing uplink spatial filter prediction on the basis of a downlink measurement result, and/or the terminal device can support performing downlink spatial filter prediction on the basis of an uplink measurement result, thereby reducing the overheads and time delay of uplink spatial filter management and/or downlink spatial filter management. A wireless communication method comprises: a terminal device sends first capability information, the first capability information being used for indicating whether the terminal device supports performing uplink spatial filter prediction on the basis of a downlink measurement result, and/or the first capability information being used for indicating whether the terminal device supports performing downlink spatial filter prediction on the basis of an uplink measurement result.