R1-2509117
discussion
Evaluation assumptions for 6GR
From Ericsson
Summary
This Ericsson Tdoc for 3GPP RAN1#123 provides comprehensive evaluation assumptions for the 6G Radio (6GR) study item, covering system simulations, antennas, channel models, traffic modeling, link budgets, and device/PA modeling. The document contains 14 numbered proposals (Proposals 1-9 with sub-proposals) and 2 observations, totaling 16 key action points.
Position
Ericsson proposes using 38.901 channel models as baseline for 6G evaluations with only minor enhancements for sensing and high-speed Doppler modeling. They propose extended FTP traffic models (eFTP1-option2/eFTP3-option2) with multiple UE classes, session-packet hierarchies, and per-class traffic fractions (αc) to capture mixed packet sizes, and require realistic bidirectional TCP flow modeling in system simulations by combining fixed network delay with SR+UL grant+transmission delay components. They support Candidate 2 link budget template (MaxCL from 38.913) for coverage targets while proposing study adaptations to Candidate 1 (38.830 template) to improve company-to-company consistency, and require MaxCL calculation to be reflected in Candidate 1. They require 6G coverage evaluations to baseline 1Rx UE antenna and minimum device bandwidth for Idle/Inactive/initial access channels, and require UE PA backoff studies to use RF simulations/MPR evaluations beyond the waveform agenda item, with RAN4 making final RF performance conclusions.
Key proposals
- Proposal 1 (Sec 2.1): Define simulation assumptions that support FWA service operation in RAN1, using scenarios already present in 38.914 and [11] as a starting point. Defer discussion of values of mechanical and electrical tilt, defining them later according to factors including scenario and base station antenna configuration. Define simulation assumption values for both maximum power and power per MHz for base station.
- Proposal 2-1 (Sec 2.2): For around 2GHz, use (M, N, P, Mg, Ng; Mp, Np) to be (4,4,2,1,1;1,4) for Outdoor Combination 1 BS antenna modeling. For around 15GHz, use specific values in provided table (2048 elements, 256 TxRU for Outdoor Comb.1; 2048 elements, 16 TxRU for Outdoor Comb.2).
- Proposal 3 (Sec 2.3): Channel models in 38.901 should be used as the baseline for 6G evaluations. Minor enhancements (e.g., for sensing, accurate Doppler modelling for high-speed scenarios) should be considered as needed.
- Proposal 4-1-1 (Sec 2.4.1.1): Support eFTP1-option2 framework for FTP Model 1 extension with multiple UE classes, session sizes Sc, packet sizes Pc, Sc/Pc packets per session, inter-arrival time Tp, and arrival rate λc based on traffic fraction αc. Packet delay bound can be defined for at least some UE classes.
- Proposal 4-2 (Sec 2.4.2): 6G system simulation methodology should include realistic modelling of bidirectional traffic flows by considering impact of TCP slow start and TCP ACK latency on throughput. In DL system simulations, UL TCP ACK delay can be modeled by combining Component 1 (fixed Core/transport/internet network delay of 5-10ms) and Component 2 (SR+UL grant+UL transmission delay based on HARQ RTT, SR availability and TTI length).
- Proposal 4-4-1 (Sec 2.4.4): Include traffic model(s) suitable for massive communications for 6G evaluations, covering: Triggered/polled reporting, Autonomous reporting (event-driven or periodic), Remote actuation, and Firmware/software upgrade.
- Proposal 5-1 (Sec 2.5.1): Support link budget template Candidate 2 based on Maximum Coupling loss (Table 7.10.1-1: MaxCL calculation template in 38.913) for the purposes of determining coverage target(s) for 6G. No need for updating this template; it can be used as is.
- Proposal 5-2 (Sec 2.5.2): Study adaptations to the 38.830 link budget template (Candidate 1) to improve consistency of parameter selection among companies while maintaining accuracy. Limit the use of link budgets as a deciding factor in physical layer design to where link level simulations and link budgets can accurately identify the net performance benefit of design alternatives to the system.
- Proposal 6 (Sec 2.6): 6G coverage evaluation studies for FR1 should use a 1Rx UE antenna and UE bandwidth corresponding to lowest-capability device as baseline assumptions, at least for DL physical signals/channels relevant for Idle/Inactive mode and initial access.
- Proposal 7 (Sec 2.7): Additional UE PA output power backoff related studies in RAN1 beyond the waveform agenda item should also use RF simulations/RF requirements. RAN4 is informed in a timely manner of RAN1 decisions impacting features that drive RF performance, and RAN4 should make final conclusions on relative RF performance of schemes studied by RAN1.
- Proposal 8 (Sec 2.8): Use the values provided in the Table in Annex A for NTN specific evaluation assumptions (including LEO-600, LEO-1200, GEO satellite orbits, S-band/Ka-band frequencies, specific EIRP densities, satellite antenna apertures, and beam diameters).
- Proposal 9 (Sec 2.9): NR as deployed by the operators should be used as baseline for evaluations comparing 6GR performance with NR.
- Proposal 4-3 (Sec 2.4.3): For the study on traffic models for generative AI services, for video related applications, use the video traffic modeling in XR TR 38.838 as starting point, and for other applications use the extensions being studied for FTP Model 1/FTP Model 3 as starting point.
- Proposal 4-4-2 (Sec 2.4.4): Adopt the traffic models in tables 2.4.4-1 to 2.4.4-4 for 6G massive communication use cases (Network triggered reporting with DL trigger 150 bytes/UL payload 1000 bytes at 1 report/hour/UE; Autonomous reporting with periodic 1 report/hour or event-driven 5% of UEs within 1-minute window; Remote actuation with 500 bytes DL/100 bytes UL ACK at 2 commands/day/UE; Firmware/software upgrade with 2 MB DL/100 kB UL ACK at 4 upgrades/year for 10% UEs within 12-hour window).
- Proposal 5-3 (Sec 2.5.2): Reflect Maximum coupling loss (MaxCL) calculation as described in Table 7.10.1-1 in 38.913 (candidate 2 template) also in the link budget template for candidate 1.