R1-2600187
discussion
Evaluation assumption for 6GR air interface
From OPPO
Summary
OPPO's Tdoc R1-2600187 provides 9 formal proposals and 2 observations across 5 evaluation areas for the 6GR air interface study, covering antenna configurations, link budget, traffic models, and NTN evaluations. The document addresses gNB antenna configuration anomalies at 7GHz, proposes UE antenna combinations up to 4T8R for typical devices, recommends using the Rel-17 CE link budget template (Candidate 1/MPL), defines FTP-3 extension with up to 3 mixed packet sizes per cell, and outlines a framework approach for AI/ML traffic models alongside NTN link budget options.
Position
OPPO proposes updating the 7GHz outdoor gNB antenna configuration to fix inconsistent element spacing (0.8λ vs 0.5λ methodology) and impractical aperture sizes exceeding 1.5m near-field limits. They support UE antenna configurations up to 4T8R as typical common assumptions across 6G features, opposing 8- or 16-element configurations as universal baseline for handheld devices. They require using Candidate 1 (MPL) link budget template from TR 38.830 for coverage evaluation, arguing it provides more realistic deployment modeling than the simplified TR 38.913 template, and demand clarification of O2I penetration margin calculation for 7GHz. For FTP-3 traffic model extension, OPPO requires single packet size per UE with at most X=3 packet size categories mapped to field-measured data rate ranges, and requires different scaling factors for packet size versus inter-arrival time to reflect throughput contributions. On AI/ML traffic, they propose studying a general framework with modular features (PDB, jitter, packet importance) that different AI services can selectively adopt, while aligning timeline with RAN2/SA2/SA4 progress rather than rushing to define specific models. For NTN, they offer two link budget options (MCL offset from TN or CNIR metric from TR38.821) and propose reusing TN UE antenna models for L/S-band native NTN support.
Key proposals
- Proposal 1 (Sec 2.1): For 7GHz, update the gNB antenna configuration as highlighted, addressing illogical spacing methodology for outdoor combination 2 and 3 (0.8λ for 32/48 elements vs 0.5λ for 64 elements) and impractical aperture sizes like 1.64m height in Combination 3.
- Proposal 2 (Sec 2.2): For UE antenna configuration, support typical configurations up to 4T8R for common evaluation, with higher antenna configurations (e.g., 8T8R, xT16R for FWA/CPE) permitted only in partial sections like MIMO, not as common assumptions across all 6G features.
- Proposal 3 (Sec 2.3): Support Candidate 1 (MPL) for the determination of link budget template for RAN1 study, reusing the Rel-17 NR coverage enhancement template from TR 38.830, which captures antenna gain, cable/connector/combiner/body losses, and shadow fading margin.
- Proposal 4 (Sec 2.3): Considering that the coverage performance of 5G mid-band (~3.5 GHz) and around 7 GHz would be evaluated, the calculation of (27) Penetration margin (dB) for 5G mid-band (~3.5 GHz) and around 7 GHz needs to clarify, since TR 38.901 assumes 20 dB O2I loss only for <6GHz.
- Proposal 5 (Sec 2.4.1): For extension of FTP model 3 with multiple packet sizes, support single packet size per UE (Alt1), with different UE distribution for different packet sizes, emulating use cases where users focus on one application type during evaluation time (~10s).
- Proposal 6 (Sec 2.4.1): For extension of FTP model 3 with multiple packet sizes, support at most X=3 packet sizes, corresponding to categorization into small (tens of kbps), medium (hundreds of kbps), and large (thousands of kbps) data rates based on field test statistics.
- Proposal 7 (Sec 2.4.1): For extension of FTP model 3 with multiple packet sizes, support different scaling factors for Si and Ti, to reflect different contribution of packet sizes to throughput (e.g., S1=K1*S2, T1=K2*T2 with K1=k*K2 where k>1).
- Observation 1 (Sec 2.4.2): When defining a traffic model for AI services, a general framework that takes into account the impacts brought by different traffic types (text, audio, image, video, XR) and various AI services could be considered.
- Observation 2 (Sec 2.4.2): RAN1 could align its study and timeline with the research progress and outcomes from RAN2 (as well as SA2 and SA4) regarding AI services and their characteristics before defining specific traffic models.
- Proposal 8 (Sec 2.5.1): Two options could be considered for link budget analysis for 6GR NTN — Option 1 increases TN MCL value to derive target MCL for NTN, or Option 2 uses the CNIR metric calculation in TR38.821.
- Proposal 9 (Sec 2.5.2): For 6GR NTN evaluation, the parameters in Table 4 for UE antenna models could be considered for LLS and SLS, using Combination 0-3 for L/S-band native NTN, reflector antenna for Ku/Ka-band with VSAT models from TR38.821.