R1-2601805
discussion
Discussion on energy efficiency
From Spreadtrum
Spreadtrum's prior position on
10.4
at
RAN1#124
· AI-synthesized, paraphrased
verify sources →
Proposes to reduce 6G UE power model complexity by removing bandwidth configurations (5%, 400MHz) from the UE power model and eliminating the maximum schedulable PDSCH throughput ratio parameter from PDCCH-only state, arguing that 400MHz support is undetermined and potentially requires CA, making it unnecessary for evaluation. Opposes introducing a baseband-only PDSCH processing state and related scaling rule, arguing bandwidth adaptation rules already reflect PDSCH processing relaxation. For micro-sleep, proposes a simple two-point scaling rule reusing the legacy 45 power units for 100MHz and defining a single X power unit for 200MHz, explicitly rejecting a more complex scaling approach. Requires that ultra deep sleep state be confined exclusively to IoT devices, not applicable to eMBB devices.
Summary
Spreadtrum presents 9 proposals and 2 observations on UE and NW power consumption modeling for 6G energy efficiency evaluations, addressing remaining issues from previous meetings including low power modes, paging rates, processing state values, and scaling rules.
Position
Spreadtrum proposes studying whether the BS LP/EE mode is an isolated state versus a mode performed during a sleep state with additional relative power value. They require reusing UE/Group paging rates from TR 38.869 and require companies to report specific values for N and RE,REF. They propose the additional energy overhead for 2Rx scales by 1.4 times that of 1Rx (yielding a value of 21) and propose specific power values for 10MHz BW with 1Rx (12 for micro sleep, 14 for light sleep, 18 for deep/ultra-deep). They oppose introducing PDSCH-only and BB-only active power states, arguing PDSCH-only power can be scaled from PDCCH+PDSCH value and BB-only relaxation is already captured by DL bandwidth adaptation rules. They propose reusing the agreed PDCCH-only bandwidth scaling rule for other DL signals/channels such as SSB, CSI-RS, and PRS, and oppose introducing new scaling rules for BD complexity or PDCCH symbol number for PDCCH-only slots, preferring to reuse the TR 38.875 scaling rule.
Key proposals
- Proposal 1 (Sec NW power model/Low power model): Study whether LP/EE mode at BS side is an isolated state or not.
- Proposal 2 (Sec UE power model/Baseline configuration): Reuse UE/Group paging rate in TR 38.869. Companies to report the value of N and RE, REF.
- Proposal 3 (Sec Power state/EE processing state): The additional energy overhead for 2Rx can also be almost 1.4 times than 1Rx, i.e., 21.
- Proposal 4 (Sec Power state/EE processing state): For 10MHz with 1Rx, following values can be considered: 12 for micro sleep, 14 for light sleep, 18 for deep/ultra-deep.
- Observation 1 (Sec Power state/EE processing state): Remaining issue 1: whether to scale additional energy overhead.
- Observation 2 (Sec Power state/EE processing state): Remaining issue 2: Values for 10MHz BW.
- Proposal 5 (Sec Power state/New active power states): Not support PDSCH-only and BB-only state and related scaling rule.
- Proposal 6 (Sec Scaling rule/Bandwidth scaling): Reuse the agreed scaling rule for PDCCH-only for other DL signal/channel processing.
- Proposal 7 (Sec Scaling rule/PDCCH-only state): Reuse the scaling rule on BD complexity for PDCCH-only slots in TR 38.875.
- Proposal 8 (Sec Scaling rule/PDCCH-only state): Not support scaling on PDCCH symbol number for PDCCH-only slots.
- Proposal 9 (Sec Scaling rule/Reference configuration for <=20MHz UE): Reuse reference configuration and relative power consumption value in TR 38.875 for <=20MHz device.