R1-2600139
discussion
Channel coding for 6GR air interface
From Huawei
Summary
This Huawei/HiSilicon contribution to the 3GPP RAN1#124 meeting discusses channel coding for the 6G radio interface, presenting 14 proposals and 57 observations across data and control channel topics. The document focuses heavily on extensions to 5G NR LDPC codes for higher data rates and provides detailed views on maintaining backward compatibility, improving area efficiency, and reusing NR Polar codes for control information.
Position
Huawei proposes a 6G LDPC extension design based on single-edge quasi-cyclic (QC) structure with enhanced BG3 using Kb=44 information columns, hybrid row-splitting-and-extension for rate-adaptive density optimization, and no puncturing in the information column to enable faster convergence within 3-5 iterations. They oppose increasing lifting size beyond Z=384, instead promoting multi-block-parallel decoding which achieves 20% higher area efficiency than single-block-parallel decoding with doubled lifting size. They propose a detailed area efficiency evaluation methodology using throughput/area calculations and require that 6G DCI size will be moderate (up to 200 bits) with scalable DCRC interleaver generation based on a provable theorem for arbitrary Kmax. They support NR-compatible polar codes for 3-11 bit UCI payloads to replace RM codes, offering >1dB gain with DMRS-less non-coherent detection and eliminating RM decoding hardware, while opposing multi-edge BGs as incompatible with 5G-LDPC decoders and rejecting non-systematic outer coding (PAC) for polar codes due to prohibitive memory and online Gaussian elimination complexity.
Key proposals
- Proposal 1 (Sec 2.2): If the necessity of LDPC extension is justified for data rate beyond NR range for high throughput scenarios, LDPC extensions supporting high code rates—e.g., up to 0.95—should be evaluated.
- Proposal 6 (Sec 2.2.6): Do not consider a lifting value larger than 384 for LDPC extension considering the additional area overhead without any performance gain under the same area efficiency.
- Proposal 7 (Sec 2.2.8): The desired design principles for BG3 are outlined below, if the necessity of LDPC extension is justified for data rate beyond NR range: Single-edge quasi-cyclic structure; Single nested BG across code rates (i.e., the same Kb), to preserve NR-like flexibility; High throughput with low complexity and improved area efficiency; Improved BLER performance under same complexity within the target code rates range.
- Proposal 8 (Sec 2.3.1): Use the following methodology to evaluate area efficiency of potential LDPC extension candidates (detailed formulas for throughput and area estimation).
- Proposal 9 (Sec 2.4.1): In high-throughput LDPC design, employ hybrid splitting and extension to match the optimal BG density within the target code rates range for high throughput scenario.
- Proposal 10 (Sec 2.5.2): If the necessity of LDPC extension is justified for data rate beyond NR range, study the following techniques for BG3: Row splitting for reduced decoding complexity and higher orthogonality/parallelism; Hybrid splitting and extension to match optimal BG density; Shifting value design for newly-split rows to maintain a single-edge quasi-cyclic structure.
- Proposal 11 (Sec 3): The necessity and motivation of polar code enhancements need to be strongly justified first by identifying the actual source of existing problems and the specific requirements that differentiate 6G from 5G control channels, rather than immediately working on detailed schemes.
- Proposal 12 (Sec 3.1): Confirm the working assumption with revision to reuse NR polar codes for control information within NR range (larger than 11 bits).
- Proposal 13 (Sec 3.2.2): If the maximum UCI payload size would exceed 1706 bits, adopt the proposed multi-CB segmentation solutions for UCI payload larger than 1706bits.
- Proposal 14 (Sec 3.3): If justifiable drawbacks of 5G RM code are identified, 6GR shall study NR-compatible polar code design for the payload sizes in the range from 3bits to 11bits for uplink control information.
- Proposal 4 (Sec 2.2.3): If the necessity of LDPC extension is justified for data rate beyond NR range for high throughput scenarios, the BG with no puncturing in the information column should be considered.
- Proposal 5 (Sec 2.2.5): If the necessity of LDPC extension is justified for data rate beyond NR range for high throughput scenarios, the new BG with increased Kb should be considered.
- Proposal 3 (Sec 2.2.1): If the necessity of LDPC extension is justified for data rate beyond NR range for high throughput scenarios, LDPC extensions capable of faster convergence—e.g., within 10 iterations—should be evaluated.