R1-2601827
discussion
On remaining aspects of waveform for 6GR
From Nokia
Nokia's prior position on
10.2.1
at
RAN1#124
· AI-synthesized, paraphrased
verify sources →
Proposes CP-OFDM and DFT-s-OFDM as defined in 5G NR as the supported baseline communication waveforms for 6G downlink and uplink respectively, with CP-OFDM for all layers and DFT-s-OFDM restricted to single-layer UL. Presents a technical case against DL DFT-s-OFDM, arguing its PAPR advantage is neutralized by transparent PAPR reduction techniques already in deployed base stations, while introducing significant limitations including contiguous frequency allocation requirements, reduced SU/MU-MIMO precoding flexibility, and multi-RAT/MRSS incompatibility in FR1. Opposes studying Zak-OTFS, presenting simulation evidence that CP-OFDM outperforms it with realistic channel estimation and asserting its claimed benefits apply only to atypical propagation conditions. For UL coverage enhancement, supports FDSS and FDSS-SE with transparent filtering, proposes PC2 high power class as baseline, and requires DWS via DCI from the first 6G release.
Summary
This document from Nokia presents 16 proposals and 11 observations on 6G waveform design, arguing for CP-OFDM and DFT-s-OFDM as baseline 6G waveforms based on 5G NR definitions. Nokia strongly opposes Zak-OTFS as a new waveform, advocates for Frequency Domain Spectrum Shaping with Spectral Extension (FDSS-SE) for low PAPR uplink, and recommends restricting multi-layer DFT-s-OFDM to single-layer transmission only.
Position
Nokia proposes that CP-OFDM and DFT-s-OFDM waveforms as defined in 5G NR are supported as the baseline communication waveforms for 6G downlink and uplink respectively, with enhancements studied only as additions rather than replacements. They present a technical case against Zak-OTFS using their own simulation results showing CP-OFDM outperforms Zak-OTFS by approximately 11 dB under realistic channel estimation and practical deployment parameters, and argue Zak-OTFS introduces unnecessary complexity through additional transforms, DD-domain buffering latency, and new receiver designs. For low PAPR uplink coverage extension, Nokia supports FDSS and FDSS-SE with transparent filtering and opposes Option 2 for subcarrier allocation where A would be defined as 2x3y5z subcarriers due to non-valid DFT size introduction. On multi-layer uplink, Nokia opposes extending DFT-s-OFDM beyond single-layer transmission for fully coherent and partially coherent UEs, citing system-level simulation results showing up to 16% cell average throughput loss when DFT-s-OFDM is used as the sole UL waveform with restricted codebook subsets, and argues the power gain from low PAPR cannot be exploited by non-power limited UEs (PH>0).
Key proposals
- Proposal 1 (Sec 2.1): CP-OFDM waveform as defined in 5G NR is supported for communications in 6G downlink, with enhancements/modifications studied as potential additions.
- Proposal 3 (Sec 2.2): RAN1 to deprioritize studying Zak-OTFS because CP-OFDM outperforms it with realistic assumptions, it provides benefit mainly in non-typical propagation conditions, and introduces major change and additional complexity.
- Proposal 6 (Sec 2.3): Frequency Domain Spectrum shaping (FDSS) and FDSS with spectrum extension (FDSS-SE) are supported in 6G Radio.
- Proposal 4 (Sec 2.3): Do not support Option 2 where the number of subcarriers A before extension/truncation is 2x3y5z subcarriers.
- Proposal 5 (Sec 2.3): Transparent filtering approach is assumed for FDSS and FDSS-SE in 6G Radio, meaning the receiver does not need to be aware of the used filter parameters.
- Proposal 9 (Sec 2.3): Dynamic waveform switching using DCI is introduced to 6G in the first release.
- Proposal 10 (Sec 2.4): Low PAPR waveform study for multi-layers UL transmission in 6G needs to consider at least fully coherent and partially coherent UEs.
- Proposal 11 (Sec 2.4): RAN1 to consider a baseline reference scenario for multi-layer waveforms study including 5G NR UL MIMO codebooks, CP-OFDM for all layers and DFT-s-OFDM for single layer, fully coherent UE, DWS rel-18 enabled, and at least rel-16 Full power modes.
- Proposal 12 (Sec 2.4): RAN1 to study DFT-s-OFDM and assess its potential from UL MU-MIMO system perspective while ensuring more flexible scheduling for paired UE and better co-existence with CP-OFDM UL waveform compared to 5G NR.
- Proposal 13 (Sec 2.4): Support CP-OFDM in uplink for all number of layers and all UE coherence capabilities, at least for non-power limited UEs where DFT-s-OFDM potential power gain cannot be exploited.
- Proposal 14 (Sec 2.4): Support DFT-s-OFDM in uplink for single layer transmission due to its advantage for power limited UEs; the potential of UL DFT-s-OFDM for more than 1 layer in 6G needs further justification.
- Proposal 15 (Sec 2.4): For 6GR, a UE supporting UL DFT-s-OFDM shall also support UL CP-OFDM for the same number of layers; only non-coherent precoder is supported for UL DFT-s-OFDM; DWS is supported.
- Proposal 16 (Sec 2.5): Adopt DFT segmentation to enhance UL DFT-s-OFDM scheduling flexibility.
- Proposal 7 (Sec 2.3): High power class should be the baseline for 6G due to significant enhancement in coverage.
- Proposal 8 (Sec 2.3): Power boosting features such as the ones specified in NR should be part of the baseline for 6G.