FS_6G_Radio

Companies present high-level visions for the 6G Radio (6GR) air interface, focusing on scalable device design paradigms, minimum spectrum allocations, multi-RAT spectrum sharing (MRSS) with 5G NR, native TN-NTN harmonization, coverage evaluation, bandwidth part (BWP) adaptation, and spectral aggregation frameworks. Contributions debate the smallest maximum UE bandwidth for low-tier IoT devices (3 MHz vs 5 MHz vs 20 MHz), propose single-cell multi-carrier (SCMC) concepts to aggregate fragmented spectrum, and recommend design philosophies ranging from bottom-up IoT-first approaches to lean designs that explicitly limit configuration flexibility learned from 5G NR. Key technical disagreements include whether to remove center frequency alignment requirements for DL/UL BWP in unpaired spectrum, what coverage evaluation metric to use (MPL vs MCL vs MIL), and how to structure initial access signals including SSB periodicity and bandwidth.

This sub-topic addresses evaluation assumptions for 6G Radio air interface studies, covering antenna modelling for BS and UE across multiple frequency bands, new traffic models for AI/ML and immersive communications, link budget methodology, NTN deployment parameters, and ISAC performance metrics. Companies are presenting proposals on whether to extend FTP models or define new token-based traffic models for AI/ML services, which link budget template to adopt as baseline (Candidate 1 from TR 38.830 vs. Candidate 2 from TR 38.913), and specific satellite orbit/antenna configurations for NTN evaluations.

This sub-topic addresses the selection and enhancement of waveforms for the 6G radio air interface. Companies are debating whether to adopt DFT-s-OFDM as an additional downlink waveform, with many opposing it due to limited PAPR gains and multiplexing challenges, while discussing uplink PAPR reduction techniques like FDSS with spectrum extension and I/Q-offset DFT-s-OFDM. Evaluation methodology, specifically the 'Net Gain' metric, and the framework for multi-layer DFT-s-OFDM in the uplink are also key areas of technical discussion.

Companies are discussing the 6G Radio (6GR) frame structure design, focusing on simplification relative to 5G NR. Key themes include reducing numerology options per band (with multiple companies proposing a single SCS per band), managing maximum UE channel bandwidth around 7GHz (with debate between ~200MHz and 400MHz single-carrier support), defining new symbol/slot types to natively support Subband Full Duplex (SBFD) rather than relying on NR Rel-19's override-based approach, and deprioritizing or removing dynamic TDD and DCI-based Slot Format Indication (SFI) in favor of semi-static RRC-configured TDD patterns. Additional discussions address FFT size limits (with opposition to 16K FFT at UE due to complexity), extended CP for sensing coverage, 7.5kHz SCS for NTN coverage enhancement, scheduling across slot boundaries, and multi-RAT spectrum sharing alignment between 5G NR and 6GR.

This sub-topic addresses channel coding enhancements for 6G data and control channels beyond 5G NR capabilities. Companies discuss LDPC extensions for higher data channel throughput (targeting 2-8x NR peak data rates) through options including fast-convergence base graph designs, increased lifting sizes, inter-CB coding, and parallelism optimizations. For control channels, discussions focus on Polar code extensions for larger DCI/UCI payloads beyond the 140-bit and 1706-bit NR limits via segmentation enhancements, interleaver redefinition, PAC codes, and high-order modulation support, alongside evaluation methodology harmonization covering metrics, assumptions, and templates.

This sub-topic covers potential enhancements to modulation and joint channel coding/modulation for 6G Radio beyond 5G NR uniform QAM. Companies are debating the feasibility and value of higher-order modulations (4096QAM DL, 1024QAM UL), with several presenting technical cases against them based on RF impairment sensitivity and EVM requirements. There is sharp disagreement on constellation shaping: Huawei, OPPO, and MediaTek present detailed evidence against probabilistic shaping (PAS/CCDM) due to performance losses in fading channels and prohibitive complexity, while supporting alternatives like enhanced AMC or enumerative sphere shaping; Ericsson, LG, and Xiaomi show more openness to geometric shaping (NUC) but with varying degrees of caution; AT&T proposes focused study of both geometric and probabilistic shaping for high-order modulations within existing NR architecture.

Companies are contributing technical proposals for 6G energy efficiency covering evaluation methodology (BS/UE power models, metrics), network-side techniques (SSB periodicity extension, on-demand SSB, cell DTX/DRX, spatial-domain adaptation, low-power radio), and UE-side techniques (C-DRX enhancements, WUS designs, BWP adaptation, PDCCH monitoring reduction). Key debates include whether to adopt BS Category 1 or Category 2 as the 6G baseline power model, whether cell DTX/DRX should apply to idle/inactive UEs alongside connected UEs, and how to structure common signal designs (always-on vs. on-demand SSB, periodicity extensions up to 160ms, simplified/light SSB concepts) to achieve native energy savings.

This sub-topic addresses AI/ML integration into the 6G radio interface, with companies proposing and debating use cases across CSI overhead reduction, DMRS overhead reduction, beam management extensions, and the AI/ML life cycle management framework. Key technical themes include neural receivers for sparse or superimposed DMRS, two-sided vs. one-sided models for CSI compression and prediction, cross-frequency and inter-cell beam prediction, and whether to build on the 5G LCM framework or design a new unified framework. Companies differ on use case selection methodology, evaluation metrics (intermediate KPIs like SGCS vs. final system performance), and the appropriate scope and timing of framework discussions.