R1-2601865 discussion

Discussion on downlink transmission scheme(s) for downlink shared channels

From ZTE
Status: not treated
WI: FS_6G_Radio
Agenda: 10.5.2.2
Release: Rel-20
Source: 3gpp.org ↗

Summary

This ZTE contribution to 3GPP RAN1#124 bis contains 33 proposals and 29 observations covering the 6GR downlink shared channel framework. It argues for prioritizing multi-TRP operation, unified DMRS design with up to 48 orthogonal ports, flexible codeword-to-layer mapping, and deprioritizing AI-based receiver and SIP studies at the beginning of 6GR.

Position

ZTE proposes that 6GR downlink MIMO framework study focus on multi-TRP operation from initial access, with CJT and sDCI-based mTRP prioritized for day 1 deployment to address coverage, UPT, and peak data rate. They require a unified DMRS design based on Comb-3 CDM groups with FD-OCC length=4 per RB per symbol, supporting up to 48 orthogonal DMRS ports plus non-orthogonal DMRS via different sequences, and explicitly deprioritize interleaved VRB-to-PRB mapping over the entire bandwidth part. For codeword-to-layer mapping, they propose studying three flexible options even for up to 4 layers, with particular emphasis on Option 2 (single CW with individual modulation order per layer) showing 36% SE gain. They present technical case against AI-based receiver for DMRS-based channel estimation, arguing the performance gain is trivial (0.5-1.5 dB) and may be eliminated by AI/ML model flaws, and deprioritize SIP study due to suboptimal performance compared to AI-based LD-DMRS with far higher computation complexity (31G vs 1.3G FLOPs).

Key proposals

Your notes

Private to your account