R1-2508983
discussion
FL summary#3 for 6G channel coding
From ZTE
Summary
This ZTE/Apple moderator summary from RAN1#123 captures ~25 proposals and 10 observations on 6G channel coding for both data (LDPC) and control (Polar) channels, focusing on extensions beyond NR range. Meta-proposals define evaluation metrics, assumptions, and templates, while technical options address LDPC throughput enhancements (e.g., fast convergence, larger lifting sizes) and Polar code extensions for larger DCI/UCI payloads (e.g., more segments, new interleavers, PAC codes).
Position
ZTE, as moderator, summarizes company positions without taking an explicit company stance; the document frames FL proposals for evaluation harmonization. They propose evaluation metrics for LDPC requiring BLER performance, complexity, and decoding throughput/latency, with area efficiency and energy efficiency as additional optional metrics. They propose LDPC evaluation assumptions specifying AWGN channel, uniform QAM modulations including at least 256QAM/1024QAM, target BLER of 10^-2 and 10^-4, and Layered BP decoding with 2-20 iterations and reversed decoding order. For LDPC extension options, they list six options including fast convergence LDPC codes, increasing maximum lifting size to [2 or 4]*384, optimizing parallelism through row orthogonality, increasing systematic columns, reducing edges in BG, and implementation-based solutions, all combinable. They propose studying error floor optimization and low complexity/power scenario enhancements for LDPC. For control channel coding, they document Polar code extension options for DCI beyond 140 bits (segmentation, D-CRC interleaver removal/extension, PAC code, 2-stage DCI, 1024-length Polar sequence, higher modulation) and UCI beyond 1706 bits (more than 2 segments, PAC code, N/2 Polar sequence, higher modulation). They propose small UCI study considering 5G RM code, enhanced coding schemes, and sequence-based approaches for 3-11 bit payloads.
Key proposals
- Proposal 3.1.1-1-v2 (Sec 3.1.1 Evaluation metrics): ...at least the following metrics should be reported: BLER performance, Complexity, Decoding throughput/latency, [Area efficiency]; [Energy efficiency] can be also reported.
- Proposal 3.1.2-1-v2 (Sec 3.1.2 Evaluation assumption): ...evaluation assumptions for LDPC extension are: AWGN channel, uniform QAM modulations (at least 256QAM/1024QAM) and code rates in MCS table, info block size reported by company, target BLER=10^-2, 10^-4, Layered BP decoding, max iterations 2~20, reversed decoding order.
- Proposal 3.2-1-v2 (Sec 3.2 Data channel coding scheme): ...for LDPC extension for higher data rate... options considered for further down-selection: Option 1: Reduce max iterations (fast convergence LDPC), Option 2: Increase max lifting size to [2 or 4]*384, Option 3: Optimize parallelism (improve orthogonality), Option 4: Increase systematic columns, Option 5: Reduce edges in BG, Option 6: Implementation based solutions.
- Proposed conclusion 3.2-2-v2 (Sec 3.2 Data channel coding scheme): ...companies are encouraged to provide LDPC BG(s) and PCM(s) in excel, provide required SNR for target BLER using template, and definition of complexity will be further discussed.
- Proposal 3.2-3-v1 (Sec 3.2 Data channel coding scheme): Study whether/how to optimize the error floor performance for LDPC code.
- Proposal 3.2-4-v1 (Sec 3.2 Data channel coding scheme): Study whether/how to enhance LDPC for 6G low complexity/power scenarios.
- Proposal 3.3-1-v1 (Sec 3.3 Data channel coding chain): ...proponent encouraged to provide details/expected spec impact of data channel coding chain enhancements (including inter-code block coding, enhancement of bit interleaver), evaluation metrics (BLER/throughput), and evaluation assumptions (channel type, resource allocation, MCS, channel estimation, antenna config).
- Observation 4.1-1-v1 (Sec 4.1 Channel coding on DCI): [16 sources] suggested using NR Polar code within NR range; [4 sources] discussed early termination; [2 sources] discussed RNTI FAR; [2 sources] discussed higher modulation order; [1 source] discussed CRC overhead.
- Observation 4.1-2-v1 (Sec 4.1 Channel coding on DCI): ...for DCI payload size >140 bits: [3 sources] suggested segmentation, [3] remove D-CRC interleaver, [4] define extended D-CRC interleaver, [2] PAC code, [1] 2-stage DCI, [2] 1024-length Polar sequence, [2] higher modulation order.
- Observation 4.2-1-v1 (Sec 4.2 Channel coding on UCI): ...within NR range: [16 sources] suggested NR Polar code; [3 sources] observed BLER degradation with NR segmentation; [3 sources] suggested new segmentation (>2 segments/new rule) with up to 2dB gain.
- Observation 4.2-2-v1 (Sec 4.2 Channel coding on UCI): ...beyond NR range (>1706 bits): [9 sources] suggested >2 segments (2~5dB gain observed for 1706-2026 bits); [1 source] suggested TPAC code; [1 source] suggested Polar encoding based on N/2 sequence (0.2dB gain).
- Proposal 4.3-1-v2 (Sec 4.3 Small code block): ...for 6G small UCI channel coding (payload <12 bits), study: use of 5G RM code, Enhanced coding scheme (including rate matching), [Use of sequence]; FFS: range of small UCI, e.g., 3~11 bits.
- Proposal 4.5-1-v1 (Sec 4.5 Control coding chain): ...study of 6G control channel coding chain enhancements, proponent to provide details/spec impact, evaluation metrics (BLER, throughput), evaluation assumptions (channel type, resource allocation, MCS, channel estimation, antenna config).
- Proposal 4.6.1-1-v1 (Sec 4.6.1 Evaluation metrics control): ...for evaluation of 6G control channel coding, at least: BLER performance, FAR, Complexity, Early termination, DCI RNTI false alarm are considered.