R1-2601815 discussion

Discussion on aspects of uplink-based CSI acquisition for 6GR

From Spreadtrum
Status: not treated
WI: FS_6G_Radio
Agenda: 10.5.3.2
Release: Rel-20
Source: 3gpp.org ↗
Spreadtrum's prior position on 10.5.3.2 at RAN1#124 · AI-synthesized, paraphrased
verify sources →
Proposes limiting 6GR Day1 SRS study to {1,2,3,4} antenna ports, noting 8-port devices lack commercial deployment despite NR support. Recommends reusing Zadoff-Chu sequences as the SRS baseline and studying cross-slot SRS features beyond 14 symbols for both CSI acquisition and velocity estimation in sensing. For AI-based SRS, asserts that antenna switching enhancement should remain NW implementation without spec impact (Observation-2), while prioritizing frequency hopping overhead reduction with Sub-case A where the NW-side model predicts full-bandwidth channel information from sparse SRS transmissions. Prefers a unified uplink RS design aggregating communication, positioning, and sensing functions, and proposes studying simplified single-step offset indication for aperiodic SRS timing determination.

Summary

This Spreadtrum contribution presents 14 proposals and 4 observations on 6G SRS design, addressing antenna port configurations, transmission schemes, AI-based overhead reduction, and integration with sensing. The document covers baseline SRS features inherited from NR alongside new study areas for 6GR including wideband transmission, cross-slot operation, and ZP SRS for CJT interference mitigation.

Position

Spreadtrum recommends focusing 6GR Day1 SRS design on {1,2,3,4} antenna ports rather than >8 ports, citing lack of commercial 8-port device deployment despite NR Rel-18 support. They propose reusing NR baselines including Zadoff-Chu sequences, periodic/SP/aperiodic time-domain behavior, and the antenna switching/codebook/non-codebook/BM usage set. They present technical case for studying TD-OCC for capacity enhancement, ZP SRS for CJT interference mitigation via reserved non-transmitted REs in PUSCH, and unified uplink RS design for integrated communication and sensing. On AI/ML, they argue AI-based SRS antenna switching is purely NW implementation with no spec impact, while AI-based SRS frequency hopping overhead reduction shows 7.5-17.9% SGCS gain and requires only minor SRS configuration changes.

Key proposals

Your notes

Private to your account