R1-2600054
discussion
Discussion on downlink transmission schemes for downlink shared channels for 6GR air interface
From FUTUREWEI
Summary
Futurewei presents 9 observations and 14 proposals on 6G DL transmission schemes, advocating for 5G NR MIMO framework evolution with AI/ML integration from day-1, focusing on DMRS adaptation, hybrid antenna architectures for UMB, frequency-selective precoding, and cooperative MIMO.
Position
Futurewei proposes adopting the existing 5G NR MIMO/RS/CSI framework as a 6G baseline while requiring AI/ML integration from day-1 of the MIMO framework design. The company proposes a UE-assisted DMRS pattern selection scheme where UEs report minimum DMRS density requirements (D_F, D_T, D_J) rather than preferred patterns, claiming over 70% reporting overhead reduction compared to direct pattern reporting. Futurewei proposes studying per-subcarrier matched-filter frequency-selective precoding that simultaneously achieves finest precoding granularity and single wideband channel estimation, addressing what they identify as a fundamental limitation of 5G NR subband precoding. The document proposes cooperative MIMO schemes leveraging TDD reciprocity with DL interference probing via SRS, presenting capacity gains of 64.8% in Dense Urban and 26.9% in Urban Macro scenarios with 45Mbps XR traffic. Futurewei emphasizes hybrid antenna architectures for UMB with large antenna element counts (2048) and moderate TXRU numbers, distinguishing them from fully-digital architectures based on subarray size and dynamic analog beamforming adjustability.
Key proposals
- Proposal 1 (Sec 2): For 6G MIMO/RS/CSI, adopt 5G NR MIMO/RS/CSI framework as a starting point and study enhancements for UMB, General Performance (SE and UPT), and Fast beam acquisition for FR2.
- Proposal 2 (Sec 2.3): RAN1 to discuss high-level MIMO framework principles that enable integration of AI/ML-based functionality from day-1 of 6G, which also facilitates evaluation for AI/ML use cases.
- Proposal 3 (Sec 2.3): Identify a few AI/ML use cases to facilitate the study and assessment of AI/ML-ready MIMO framework design.
- Proposal 4 (Sec 3.2): For 6GR DMRS enhancements, support UE reports DMRS densities values to facilitate DMRS pattern selection.
- Proposal 5 (Sec 3.3): Consider including AI/ML-based DMRS configuration optimization as a candidate approach for DL DMRS enhancement in 6G study.
- Proposal 6 (Sec 3.4): For UMB base stations, hybrid antenna architecture with a large number of antenna elements and moderate number of TXRUs supporting hybrid beamforming should be a key focus area.
- Proposal 7 (Sec 3.4): For UMB base stations, at least study specific hybrid antenna combinations: for 7 GHz focus on 2048 elements/256 TXRUs and fewer TXRUs; for 15 GHz focus on 2048 elements/32 TXRUs and 2048/128 TXRUs.
- Proposal 8 (Sec 3.4): Study potential MU MIMO transmission schemes for hybrid antenna architectures, with different approaches for full CSI vs. analog-beam-specific CSI availability.
- Proposal 9 (Sec 3.5): Study fine-granularity frequency-selective precoding in DL transmissions without the need of separate/multiple UE-side channel estimations, e.g., per-subcarrier matched filter precoding.
- Proposal 10 (Sec 3.6): Study cooperative MIMO via DL interference probing based on SRS enhancements to improve 6G system capacity for TDD.
- Proposal 11 (Sec 3.6): Study cooperative MIMO for TDD with downlink interference probing via SRS, associating PRB/port allocation of a PDSCH transmission with a corresponding aperiodic SRS transmission.
- Proposal 12 (Sec 3.7): Study how to achieve desired tradeoff between capacity and reliability in 6GR by utilizing many degrees of freedom in spatial/frequency/time domains, supporting transmit diversity schemes and multiplexing of diversity/spatial multiplexing schemes.
- Proposal 13 (Sec 3.8): In 6GR, consider the unified TCI framework developed in 5G NR as a starting point for 6GR QCL/TCI framework.
- Proposal 14 (Sec 3.8): In 6GR, support QCL/TCI enhancements to enable the use of UL signal (e.g., SRS) as QCL source RS of DL TCI states.