R1-2601879 discussion

Beam management for downlink and uplink

From Kyocera Corporation
Status: not treated
WI: FS_6G_Radio
Agenda: 10.5.2.4
Release: Rel-20
Source: 3gpp.org ↗
Kyocera Corporation's prior position on 10.5.2.4 at RAN1#124 · AI-synthesized, paraphrased
verify sources →
Proposes reusing the Rel-18/19 inter-cell AI/ML beam management framework for Distributed MIMO (D-MIMO) scenarios with multi-TRP transmission, using 5G NR 2-TRP operation as a baseline reference and studying m-TRP operation with more than 2 TRPs while documenting throughput gains versus complexity. Proposes studying spatial domain beam prediction to reduce the number of searches and processing time across multiple TRP beam candidates, specifically reusing the Set A / Set B mechanism from Rel-18/19 for efficient TRP selection based on UE measurement reports. Proposes extending temporal-domain beam prediction to D-MIMO to improve beam tracking and enable seamless TRP switching for mobile UEs. Proposes studying beam management for multiple TRP deployments considering multiple UEs connecting to multiple TRPs simultaneously, and specifically studying efficiency improvements for CSI-RS reporting in D-MIMO, noting that current L1-RSRP reporting limits of N = 6 or 8 may provide insufficient prediction accuracy for joint transmission scheduling.

Summary

Kyocera's R1-2601879 contribution for 6GR beam management presents 11 proposals and 8 observations covering simulation assumptions, QCL/beam indication, beam measurement/report, and UE-initiated beam management (UEIBM), advocating for a unified TCI/QCL framework baseline, simplified QCL chains, enhanced mTRP and inter-cell signaling, and a refined event-driven UEIBM procedure with conditional network response.

Position

Kyocera proposes adopting the Rel-17/18 unified TCI framework and existing QCL rules as the baseline for 6GR beam indication without redesign. They propose studying a streamlined QCL property/chain design that evaluates removing or reducing reliance on UE-specific periodic TRS as a QCL source RS, explicitly assessing impacts on tracking/channel estimation robustness, system performance, and overhead/energy consumption. They propose studying low-latency beam indication mechanisms including prediction-assisted methods realizable with AI/ML and non-AI approaches to reduce beam update latency in high-mobility and dense mTRP deployments. For AI/ML beam management, they propose studying methods to address associated ID limitations including potential disclosure of NW proprietary information across sTRP and mTRP scenarios. For UEIBM, they propose eliminating Events 1, 2, 3, and 9 as redundant, and modifying Event-5 to support mTRP dynamic switching between DPS and CJT modes where the UE reports a group of M TRPs meeting quality conditions. They propose studying and specifying an event-driven UEIBM UL reporting framework with aperiodic UL procedure and conditional network response only when beneficial.

Key proposals

Your notes

Private to your account