R1-2410588
discussion
Discussions on specification support for positioning accuracy enhancement for AI/ML
From ITL
Summary
ITL presents technical specifications for AI/ML-based positioning enhancements in NR, addressing timing reference methods and training data generation entities across multiple positioning cases. The document contains 3 main proposals covering reference time definitions for multi-RTT support and standardizing training data label generation sources.
Position
ITL advocates FOR unified reference timing approaches across different AI/ML positioning cases, supporting multi-RTT capability by using different timing references for UL and DL parts. They push FOR including multiple entities (PRU, Non-PRU UE, LMF) as valid sources for training data label generation, promoting flexibility in training data collection while maintaining consistency between training and inference phases.
Key proposals
- Proposal 1 (Sec 2): For AI/ML based positioning Case 3b, the existing UL RTOA reference time T0+tSRS is applicable for only UL part, and TRP transmit timing of DL subframe should be used to support multi-RTT
- Proposal 2 (Sec 2): For AI/ML based positioning Case 2b, option 1 (TSubframeRxi) should be used to support DL-TDOA, and option 2 (TUE-TX) should be used to support multi-RTT
- Proposal 3 (Sec 3): For training data generation of AI/ML based positioning Case 1/2a/2b/3b, the label and its related data can be generated by PRU, Non-PRU UE with estimated location, or LMF