R1-2410735
discussion
FL summary #2 for AI/ML in beam management
From Samsung
Summary
This 3GPP RAN1 technical document (Tdoc R1-2410735) from Samsung serves as FL summary #2 for AI/ML in beam management, containing over 200 proposals and observations across multiple technical areas including performance monitoring, configuration methods, inference reporting, and beam indication mechanisms.
Position
Samsung advocates for a comprehensive framework supporting both UE-side and NW-side AI/ML models with emphasis on: 1) Flexible applicability reporting combining multiple options rather than forcing single approach, 2) Dedicated monitoring configurations separate from inference to avoid complexity, 3) Practical reference time definitions based on actual transmission occasions rather than abstract timing, 4) Rich content options for data collection including multiple types to support different model requirements, and 5) Reuse of existing CSI framework mechanisms where possible to minimize specification impact.
Key proposals
- Proposal 1.1B (Sec 1.1): Direction A and Direction B for combining Options 1 and 2 for applicability reporting in UE-side model configuration
- Proposal 2.1-1a (Sec 2.1): Introduce beam accuracy indicator (BAI) for beam prediction accuracy in CSI report for monitoring with BAI = Np/N where Np is correct predictions out of N total instances
- Proposal 2.2-1a→b (Sec 2.2): Support dedicated resource set(s) for monitoring configured in dedicated CSI report configuration with CSI-ReportConfigID linking to inference configuration
- Proposal 3.1a (Sec 3.1): Support Option 3 for reference time based on latest transmission occasion of CSI-RS/SSB resource in Set B for BM-Case2
- Proposal 4.1a (Sec 4): Support inference result reporting for BM-Case 1 with beam information and RSRP of predicted Top K beams
- Proposal 5.6 (Sec 5.3): Support beam information options when M < size of measurement resource set including CRI/SSBRI or bitmap indication
- Proposal 6.1 (Sec 6.1): Support reporting measurement results of multiple time instances in one report for NW-sided model BM-Case2
- Proposal 7.1 (Sec 7.1): Study extension of Rel-17 TCI state activation/indication signaling for N future time instances
- Proposal 8.1 (Sec 8.1): Existing CPU mechanism used as starting point for AI/ML-based CSI processing with FFS on sharing between legacy and AI/ML reporting
- Proposal A (Sec 9): Direction A supporting both CSI-ReportConfig and inference related parameters for applicability reporting
- Proposal B (Sec 9): Reuse CSI framework for monitoring Type 1 Option 2 configuration with dedicated resource sets
- Proposal C (Sec 9): Support three types of contents for data collection - Type 1: L1-RSRPs only, Type 2: L1-RSRPs + beam info, Type 3: L1-RSRPs + beam info + RSRPs
- Proposal D (Sec 9): Ranking information of predicted Top K beams conveyed by order of beam information based on model output