R1-2601804
discussion
Discussion on modulation for 6GR
From Spreadtrum
Spreadtrum's prior position on
10.3.2
at
RAN1#124
· AI-synthesized, paraphrased
verify sources →
Opposes the adoption of Probabilistic Shaping (PS) for 6G, arguing that its serial and block-based Distribution Matcher/De-matcher (DM/DDM) processing fundamentally conflicts with the TS 38.214 PDSCH processing time (N1) constraint, leading to prohibitive chip area and buffer costs. Presents a technical case against supporting joint channel coding and modulation in 6G Release, though does not preclude discussing its use case in 6G AI contexts. Questions the statistical significance of DL 4096QAM and UL 1024QAM, citing system-level simulations showing that less than 3% of sub-7GHz UMa UEs and only 10-23% of mmWave UEs achieve the required SIR/SINR thresholds. Requires that any evaluation of shaping gains be based on a Net Gain methodology that subtracts PAPR-induced PA backoff penalties and non-ideal implementation losses from theoretical AWGN gains.
Summary
In R1-2601804, Spreadtrum presents discussion and proposals on 6GR modulation across four areas: higher-order modulation evaluation, constellation shaping (probabilistic and geometric), PAPR evaluation, and control channel modulation. The document contains 11 observations and 8 proposals, arguing against probabilistic shaping due to structural conflicts and supporting geometric shaping with down-selection and independent parallel evaluations for higher-order modulation.
Position
Spreadtrum proposes parallel and independent evaluations between RAN1 and RAN4 for DL 4096QAM and UL 1024QAM, with RAN1 focusing only on ideal baseband performance without RF impairment modeling, and requires alignment with RAN4's existing prioritization of UL 1024QAM. They present a technical case against probabilistic shaping (PS) adoption in 6GR, arguing that the fundamental principle of PS causes structural conflicts with legacy NR payload-independent processing including scrambling, interleaving, and CB segmentation, and inevitably requires LDPC redesign due to incompatibility with legacy filler bits. They support geometric shaping (GS) as it remains completely transparent to all other Tx/Rx chain functionality blocks except mapper and modulation blocks, requiring only LUT updates for constellation and bit mapping rules, and propose down-selection between 1D-NUC and 2D-NUC during the RAN1/RAN4 Malta joint session. They require quantifiable PAPR increase values to be provided alongside performance gains and jointly evaluated with performance-complexity trade-offs to determine shaping feasibility. They support legacy 5G NR equal probability uniform QPSK for the 6GR control channel, citing mandatory robustness requirements due to the absence of HARQ retransmission for DCI/UCI payload.
Key proposals
- Proposal 1 (Sec 2.1): For the evaluations of DL 4096QAM and UL 1024QAM, RAN1 and RAN4 should conduct parallel and independent evaluations. Specifically, RAN1 focuses on ideal baseband performance evaluation without modeling RF impairments.
- Proposal 2 (Sec 2.1): RAN1 should align with RAN4 to prioritize the evaluation of UL 1024QAM.
- Proposal 3 (Sec 2.1): RAN1 and RAN4 should timely exchange respective evaluation progress to keep each other informed.
- Proposal 1 (Sec 2.2.1): PS modulation is not considered for 6GR.
- Proposal 1 (Sec 2.2.2): In order to maintain a reasonable workload on studying GS for both RAN1&RAN4, it is proposed to down-select the GS schemes (e.g., 1D-NUC vs 2D-NUC) based on the performance and complexity trade-offs during the RAN1 and RAN4 joint session at Malta meeting.
- Proposal 1 (Sec 2.3): It is proposed to provide the corresponding increased PAPR value together with the performance gains.
- Proposal 2 (Sec 2.3): The quantifiable PAPR increased and the performance-complexity trade-off should be considered together as factors of feasibility.
- Proposal 1 (Sec 2.4): 5G NR equal probability uniform QPSK is supported for 6GR control channel. Note: Whether to use other modulation schemes for the 6GR control channel than QPSK modulation is separately discussed.