R1-2600029
discussion
On remaining aspects of modulation in 6GR
From Nokia
Summary
Nokia provides 11 proposals and 5 observations on 6G modulation, establishing baseline uniform QAM schemes from 5G NR (with UL 1024QAM added) and addressing enhancements like geometric/probabilistic constellation shaping, higher-order QAM, and MCS table optimization.
Position
Nokia proposes adopting 5G NR uniform QAM schemes as baseline 6G modulation and adding uniform 1024QAM in UL. They oppose uniform 4096QAM in DL due to R-ML MIMO detection complexity and EVM requirements. For geometric constellation shaping, they require down-selection from higher-order uniform QAM with sign symmetry, quantization with limited bits per I/Q, and minimization of shape variants per order. They deprioritize AI two-sided model approaches while allowing AI receiver evaluation. They present a technical case against probabilistic constellation shaping, arguing AWGN gains do not carry over to fading channels and highlighting impacts on HARQ retransmissions, PAPR, and MIMO detector complexity. They propose studying MCS tables that allow multiple modulation-order/code-rate pairs per spectral efficiency point.
Key proposals
- Proposal 1 (Sec 2): For 6GR DL, 5G NR uniform QPSK, 16QAM, 64QAM, 256QAM and 1024QAM are supported for data channel; for UL, QPSK, 16QAM, 64QAM, 256QAM (CP-OFDM) and pi/2 BPSK, QPSK, 16QAM, 64QAM, 256QAM (DFT-s-OFDM).
- Proposal 2 (Sec 3.1): Support uniform 1024QAM in Uplink.
- Proposal 3 (Sec 3.2): Address the specification impact of geometric shaping and probabilistic shaping, e.g. on DL-MU-MIMO.
- Proposal 4 (Sec 3.2.2): Geometrically shaped constellations shall be a down-selection (down-sampling) of a higher order uniform QAM. Sign symmetry per I/Q shall be maintained.
- Proposal 5 (Sec 3.2.2): Geometrically shaped constellations shall be quantized based on a higher order uniform QAM constellation. To improve the quantization error, the even amplitudes between two legible odd amplitudes are considered as legible quantization result. Sign symmetry per I/Q shall be maintained.
- Proposal 6 (Sec 3.2.2): Evaluation of geometrically shaped constellations shall take the quantization into account. The quantization can be considered by assuming a maximum number of bits, e.g. 4, 6 or 8 bits per I/Q.
- Proposal 7 (Sec 3.2.2): Strive to reduce the number of different geometrically shaped constellations of the same order.
- Proposal 8 (Sec 3.2.3): For the evaluation of geometrically shaped constellations, deprioritize the study on AI two-sided model approach for constellation shaping.
- Proposal 9 (Sec 3.2.3): For the evaluation of geometrically shaped constellations, consider AI receiver as potential receiver assumption (in addition to agreed assumptions on LMMSE (UL), rML/LMMSE (DL)), where geometrically shaped constellations can be optimized for AI receiver.
- Proposal 10 (Sec 3.2.4): The trade-off between the potential gain of PCS techniques and their corresponding impact on the processing chain w.r.t. latency, complexity, PAPR, (sub-)optimality for retransmission should be considered.
- Proposal 11 (Sec 3.3): Study the possibility of allowing spectral efficiency operating points achievable with more than a single modulation order.