R1-2601896
discussion
Discussion on beam management for downlink and uplink
From TCL
Summary
This document from TCL presents 24 proposals and 7 observations on beam management for 6G RAN, covering NR Rel-19 evolution, AI/ML integration, initial access enhancements, multi-carrier/bandwidth scenarios, and UE-initiated mechanisms as baseline features from day one.
Position
TCL proposes retaining NR's hybrid architecture of analog antennas and digital links as a baseline for 6G millimeter wave and 7-24 GHz bands, evolving NR beam management with targeted enhancements. They propose that legacy periodic/semi-persistent beam reporting and UE-initiated beam reporting should both be supported as baseline mechanisms, and propose studying UE-initiated/event-driven beam reporting specifically in MTRP scenarios, with notification information enhanced to carry more event details beyond SR-like reuse. They propose adopting the unified TCI framework from day one while ensuring flexibility and scalability and reducing signaling overhead and latency, and propose extending the framework to serve both communication and sensing. They propose introducing UE-initiated beam failure recovery mechanisms and UE-initiated candidate beam extension to reduce latency and increase flexibility compared to network-controlled configurations. They propose studying a unified AI/ML-based framework that integrates RS configuration and measurement configuration for both BFD and BFR, with spatial domain beam prediction across inter-port, inter-TRP, and inter-cell levels to reduce RS overhead and enable proactive beam switching. They propose studying frequency-division and time-division beam scanning to reduce initial access latency as beam counts scale, and propose completing early beam refinement before early CSI acquisition, with msg2-related and msg4-related channel repetition supported for beam refinement. They propose studying a new discrete spectrum mechanism referred to as SCMC with enhanced beam management for multi-carrier, and propose studying large-bandwidth beam management around 7 GHz with differentiation between single and dual RF chain UE architectures.
Key proposals
- Proposal 1 (Sec 2.1): 6GR should continue to employ the hybrid architecture of analog antennas and digital links for millimeter wave and 7-24 GHz bands, using NR beam management as baseline.
- Proposal 1 (Sec 2.2): Legacy periodic/semi-persistent beam reporting and UE-initiated beam reporting should be supported as baseline mechanisms with further enhancements studied.
- Proposal 1 (Sec 2.3): 6GR should adopt the unified TCI framework from day one, ensuring flexibility and scalability while reducing signaling overhead and latency.
- Proposal 1 (Sec 2.4): UE-initiated beam failure mechanism should be introduced for beam failure recovery, with UE-initiated candidate beam extension studied.
- Proposal 1 (Sec 2.5.1): Study to integrate RS configuration and measurement configuration of BFD and BFR under a unified AI/ML-based framework.
- Proposal 1 (Sec 2.5.2): Study the benefits of spatial domain beam prediction, including spectral efficiency gains and mobility robustness.
- Proposal 1 (Sec 2.6.1): 6GR should study frequency-division and time-division beam scanning and SSB patterns during initial access.
- Proposal 1 (Sec 2.6.2): 6GR should study early beam refinement during initial access, to be completed before early CSI acquisition.
- Proposal 1 (Sec 2.6.3): 6GR should study AI for beam inference and prediction during initial access, with inferred beam index mapped to RO time-frequency positions and preamble grouping.
- Proposal 1 (Sec 2.7): 6GR should study new discrete spectrum mechanisms to fully unleash the discrete spectrum, such as SCMC (Single Cell with Multiple Carriers).
- Proposal 1 (Sec 2.8): 6GR should study beam management for operation with larger bandwidths around 7 GHz including both network-initiated and UE-initiated mechanisms.
- Proposal 2 (Sec 2.2): 6GR should study UE-initiated/event-driven beam reporting in the MTRP scenario.
- Proposal 3 (Sec 2.2): 6GR should study using notification information in UE-initiated/event-driven beam reporting to carry more information.
- Proposal 4 (Sec 2.6.2): Support introduction of repetition for msg2-related and msg4-related channels used for beam refinement during initial access.
- Proposal 2 (Sec 2.7): 6GR should study beam management for SCMC.