R1-2500274
discussion
Discussion on specification support for beam management
From CMCC
CMCC's prior position on
9.1.1
at
RAN1#119
· AI-synthesized, paraphrased
verify sources →
Proposes that L1 signaling be supported for NW-sided training data collection and that Top-K beam sweeping be supported for NW-side inference to enhance prediction accuracy. Requires that for UE-side models, the overall CPU be separately counted between legacy CSI reporting and AI/ML-based CSI reporting, while being shared among AI/ML features. Prefers Option 1 for BM-Case 2 differential RSRP reporting, which includes CRI of top-K predicted beams per instance. Supports dedicated resource sets and report configurations for UE-side model monitoring (Type 1 Option 2), linking the monitoring configuration to an inference report configuration via CSI-ReportConfig ID. Proposes that the granularity of UE capability reporting for AI/ML be at the sub-use case level, including details on Set A/B size, RS type, and model outputs.
Summary
CMCC presents 55 proposals and 4 observations regarding the specification impact of AI/ML-based beam management in NR, covering data collection, inference, and monitoring for both NW-side and UE-side models. The document addresses critical aspects such as Set A/B configuration, Top-K beam sweeping necessity, reporting content for inference results, and performance monitoring mechanisms.
Position
CMCC proposes supporting Type 1 and Type 3 data collection contents for NW-side training, emphasizing flexibility for classification and regression models. They require that RS associated with TCI state indication must be measured at least once before application, ensuring QCL parameter validity. For UE-side inference, they prefer Option 1 for reference time determination in BM-Case 2 and support both predicted and measured RSRP reporting options for BM-Case 1. CMCC argues that applicable functionality determination should be up to UE implementation rather than strictly dependent on associated ID matching, allowing for model generalization. They propose separate CPU counting for legacy vs. AI/ML CSI reporting to prevent performance degradation of legacy features. For monitoring, they prefer statistical beam prediction accuracy over N instances and support monitoring sets as down-sampled subsets of Set A to reduce overhead.
Key proposals
- Proposal 1 (Sec 2.1.1): Supports Type 1 (all L1-RSRPs) and Type 3 (L1-RSRPs of Set B + K beams of Set A) for NW-side training data collection via higher layer signaling.
- Proposal 4 (Sec 2.2.1): Proposes that UE can report or request gNB for preferred Set B patterns for UE-side model training data collection.
- Proposal 9 (Sec 3.1.1): Proposes supporting Top-K beam sweeping procedure for NW-side inference to improve prediction accuracy.
- Proposal 13 (Sec 3.1.3): Supports CRI/SSBRI or bitmap + one CRI/SSBRI for beam information in NW-side inference reporting when M < size of measurement resource set.
- Proposal 16 (Sec 3.1.4): Requires that the RS associated with TCI state indication should be measured at least once before TCI state indication or application.
- Proposal 20 (Sec 3.2.4): Proposes reporting inference results of N future time instances in one report for UE-side BM-Case 2, with reference time based on UL slot + offset.
- Proposal 21 (Sec 3.2.5): Supports both Option A (Predicted RSRP) and Option B (Predicted RSRP if not measured, measured L1-RSRP if measured) for UE-side BM-Case 1 inference results.
- Proposal 26 (Sec 3.2.6): States that determination of applicable functionality is up to UE instead of depending on the matching of associated ID.
- Proposal 37 (Sec 4.1): Supports Type 2 monitoring of UE-side AI/ML model with gNB indication of the UE’s decision.
- Proposal 42 (Sec 4.2): Prefers Option 2 for new quantity of beam prediction accuracy in Type 1 performance monitoring, calculating accuracy over N linked prediction instances.
- Proposal 48 (Sec 4.2): Supports monitoring set as a subset of Set A with uniform or non-uniform down-sampling to reduce measurement overhead.
- Proposal 54 (Sec 5): Proposes that overall CPU is separately counted between legacy CSI reporting and AI/ML-based CSI reporting, but shared among AI/ML features.
- Proposal 55 (Sec 6): Proposes granularity of UE capability for AI/ML at sub-use case level (BM Case 1/2) including Set A/B size, RS type, and time instance intervals.