Hierarchical Beam Training for Millimeter-Wave Systems
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Efficient beam training in wireless communication systems, particularly in millimeter-wave devices, is resource-consuming due to the need for exhaustive searches to select proper transmitting and receiving beams, especially when both TX and RX beam training are required, leading to increased complexity.
Innovation Solution
The implementation of hierarchical beam training methods, where beams with coarser resolution are trained first, and then finer beams are selected based on previous training results, using detection metrics to determine optimal beams and providing training gaps in downlink opportunities to avoid data loss during beam training.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If exhaustive beam search is performed to select optimal transmitting and receiving beams, then beam selection accuracy is improved, but training time and system complexity increase significantly
Solution Approach 1:
The patent divides the beam training process into two distinct stages: first stage beam training for initial beam pair identification, and second stage beam training for refined beam selection. This segmentation allows the system to achieve accurate beam selection without requiring a single exhaustive search, thereby reducing overall training time while maintaining beam selection accuracy.
Solution Approach 2:
The patent performs preliminary beam training in the first stage to identify a subset of candidate beam pairs before conducting the second stage training. This preliminary action narrows down the search space, allowing the system to achieve final beam selection accuracy with reduced computational complexity and time in the second stage.
2Measurement precision
If exhaustive beam search is performed to select optimal transmitting and receiving beams, then beam selection accuracy is improved, but device complexity increases
Solution Approach 1:
The patent segments the beam training into two stages with different complexity requirements. The first stage uses simpler procedures to identify candidate beam pairs, while the second stage applies more refined selection criteria only to the reduced candidate set, thereby achieving high beam selection accuracy with manageable system complexity.
Solution Approach 2:
The patent performs partial beam training in the first stage to identify a subset of promising beam pairs, rather than exhaustively evaluating all possible beam combinations. This partial action reduces the computational burden and device complexity while still enabling accurate beam selection in the subsequent second stage.
3Reliability
If training gaps are provided in downlink opportunities, then data loss during beam training is prevented, but scheduling flexibility is reduced
Solution Approach 1:
The patent schedules training gaps in advance within the downlink opportunities, allowing the system to proactively prevent data loss during beam training. This preliminary scheduling ensures that training activities do not conflict with data transmission, maintaining reliability while providing a structured framework for future scheduling decisions.
Solution Approach 2:
The patent implements a dynamic scheduling mechanism where training gaps are inserted selectively based on the specific beam training requirements and traffic conditions. This dynamic approach allows the system to maintain scheduling flexibility by adapting the placement and duration of training gaps to different operational scenarios, rather than using fixed rigid scheduling.
Data Source
AI summary
A network control device includes a wireless communications module and a controller. The wireless communications module uses a preferred transmitting beam to communicate with a communications apparatus in one or more downlink opportunities corresponding to the preferred transmitting beam. The controller schedules signal or data to be transmitted in at least one downlink opportunity corresponding to the preferred transmitting beam. When scheduling signal or data to be transmitted, the controller further provides at least one training gap, in which the controller does not schedule any dedicated data to the communications apparatus, in the downlink opportunity corresponding to the preferred transmitting beam.