Neural Beam Selection for Low-Overhead High-Mobility Networks
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing beam management methods for large-scale antenna arrays in mmWave communication face high signaling overhead and difficulty in adapting to rapid changes in high-mobility scenarios, leading to degraded system performance.
Innovation Solution
A network element and method that determines a preferred beam from multiple first beams based on measurement results, utilizing historical data and neural networks to establish correlations between first and second beams, without requiring additional measurements from the user.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If beam measurement is performed using SSB and CSI-RS signals in large-scale antenna arrays, then communication effect is improved, but signaling overhead is significantly increased
Solution Approach 1:
The patent extracts and utilizes terminal movement information (velocity vector, acceleration) as a separate feature from traditional beam measurement signals. By extracting movement characteristics, the system can infer beam status changes without requiring extensive beam measurement signaling, thus reducing overhead while maintaining communication effectiveness.
Solution Approach 2:
The patent performs preliminary beam selection by determining a first candidate beam set based on terminal movement information before actual data transmission. This preliminary action reduces the need for extensive beam measurement and reporting, as the candidate beam set is pre-determined based on predicted terminal behavior and channel conditions.
2Ease of operation
If traditional beam measurement methods are used, then beam selection is achieved, but system performance is degraded in high-mobility scenarios
Solution Approach 1:
The patent introduces dynamic beam management by continuously updating the candidate beam set based on real-time terminal movement information. The beam management system adapts dynamically to changing channel conditions caused by high mobility, rather than relying on static pre-configuration, thus maintaining system performance in high-mobility scenarios.
Solution Approach 2:
The patent implements a feedback mechanism where terminal movement information is continuously fed back to the network side. This feedback enables the network to adjust beam selection in real-time, ensuring that beam management keeps pace with rapid channel changes in high-mobility scenarios and maintains optimal system performance.
3Device complexity
If coarse-grained beam measurement using SSB is performed first, then beam search is simplified, but fine-grained beam measurement requires additional CSI-RS signaling overhead
Solution Approach 1:
The patent makes terminal movement information serve multiple functions: it is used for both coarse-grained beam search and fine-grained beam refinement. By utilizing movement information universally across different beam management stages, the system eliminates the need for separate CSI-RS-based fine-grained measurement, reducing overall signaling overhead while maintaining beam search effectiveness.
Data Source
AI summary
The present disclosure provides a beam selection method and a network element. The network element includes a first network element in a communication network, including: a receiving unit configured to receive measurement results of a terminal for first information transmitted via at least a part of first beams; a processing unit configured to determine a preferred beam from a plurality of second beams according to measurement results of the first information transmitted via more than two first beams; and a transmitting unit configured to transmit information about the preferred beam.