R1-2410150
discussion
AI/ML based Positioning
From Google
Summary
Google's contribution discusses ML-based positioning for NR, presenting 4 key proposals covering training data collection, model monitoring, and measurement configurations. The document builds on previous RAN1 agreements and focuses on Case 1 UE-based positioning with emphasis on transparent model performance monitoring.
Position
Google advocates FOR transparent model performance monitoring that does not require UEs to disclose location information to the network, supporting UE-side monitoring (Option A) over network-side monitoring (Option B). They push FOR flexible channel measurement options including phase information and configurable sample-based vs path-based reporting, while supporting reuse of existing measurement frameworks rather than creating new mechanisms.
Key proposals
- Proposal 1 (Training data collection): Support two options for channel measurement - Option 1: Time information and power information, Option 2: Time information, power information and phase information
- Proposal 2 (Training data collection): Support the L1-SINR as the quality indicator of channel measurement
- Proposal 3 (Training data collection): Support the NW to configure whether the UE should report the sample-based or path-based measurement results
- Proposal 4 (Model monitoring): The model monitoring should not require the UE to report the model inference result to the NW, with Option A preferred over Option B and transparent model performance monitoring
- Agreement (Training data collection): Quality indicator of timing information uses existing IE for timing quality (NR-TimingQuality in 37.355 and Timing Measurement Quality in 38.455)
- Agreement (Case 3b model inference): Mandatory support for channel measurement and time stamp, optional support for quality of channel measurement in measurement reports from gNB to LMF
- Agreement (Case 3a timing): Mandatory support for timing information and time stamp, optional support for quality of timing information using existing Timing Measurement Quality IE
- Agreement (Sample-based measurement): Defined as Nt' samples with timing granularity T=2k×Tc where k is timing reporting granularity factor and Tc is basic NR time unit
- Agreement (Path-based measurement): Refers to existing Rel-18 measurements with potential enhancements on number of reported paths
- Agreement (Case 1 assistance data): Study whether existing UE-based DL-TDOA and DL-AoD assistance data should be explicitly indicated, implicitly indicated, or other for training/inference consistency