R1-2410716
discussion
Summary #3 of specification support for positioning accuracy enhancement
From Ericsson
Ericsson's prior position on
9.1.2
at
RAN1#118bis
· AI-synthesized, paraphrased
verify sources →
Advocates for sample-based measurements over legacy path-based measurements for superior performance and lower complexity, while strongly opposing phase information inclusion in model inputs due to deployment costs and minimal accuracy gains.
Summary
This document is RAN1#119 Summary #3 from Ericsson covering AI/ML positioning accuracy enhancement with approximately 120+ proposals/conclusions across model input, output, training data collection, inference, and monitoring. The document extensively discusses sample-based vs path-based measurements, LOS/NLOS indicators, and assistance data for ensuring consistency between training and inference.
Position
Ericsson as document moderator presents a balanced compromise approach, advocating FOR: (1) supporting both sample-based and path-based measurements to avoid blocking progress despite company divisions, (2) reusing existing signaling frameworks and IEs where possible to minimize specification impact, (3) label-free monitoring methods for implementation flexibility, and (4) explicit provision of assistance data rather than implicit associated IDs. They push AGAINST: (1) supporting CIR/phase information due to lack of consensus, (2) independent LOS/NLOS reporting without associated measurements, and (3) overly restrictive definitions that limit implementation freedom.
Key proposals
- Proposal 2.1.1.2-1 (Sec 2.1): For sample-based measurement, starting time of Nt consecutive samples is timing of first detected sample within a search window determined by configured offset relative to reference time, reusing existing 'Search Window Information' IE in 38.455
- Proposal 2.1.1.2-2 (Sec 2.1): Candidate values for Nt include at least 16, 32, 64, 128; for Nt' include at least 9, 16, 32; for k include at least 0…5
- Proposal 2.1.3.2-1B (Sec 2.1): For Rel-19 AI/ML based positioning, support Alternative (a) sample-based measurement report (majority view)
- Conclusion 2.4.2-1 (Sec 2.4): For Rel-19 AI/ML based positioning for all use cases, there is no consensus in RAN1 to support CIR for determining model input
- Proposal 3.1.2-1A (Sec 3.1): For AI/ML assisted positioning Case 3a, LOS/NLOS indicator reuses meaning and format of existing IE 'LoS/NloS Information' in 38.455 when associated measurement obtained by legacy methods
- Proposal 3.1.4-1 (Sec 3.1): For measurement report of AI/ML assisted positioning Case 3a, indicate to LMF whether timing information is obtained by legacy method or by AI/ML
- Proposal 4.1.4-1 (Sec 4.1): For AI/ML based positioning Case 1, if channel measurement/label reported by LMF to UE, include both NR-TimeStamp and UTCTime from TS 37.355, with UTCTime as optional field
- Proposal 4.4.2-1 (Sec 4.4): If Part A and Part B generated by same entity, provide paired Part A and Part B in one report; paired entries share same time stamp
- Proposal 5.1.2-1A (Sec 5.1): For AI/ML based positioning Case 1, if necessary, provide assistance information from LMF to UE explicitly, as optional fields per existing specification
- Proposal 6.1.2-1 (Sec 6.1): For AI/ML positioning Cases, support label-free monitoring methods where model inference entity performs self-monitoring without external ground truth information
- Proposal 6.2.2-1A (Sec 6.2): For model performance monitoring of AI/ML positioning Case 1, support at least Option A-1 and A-2 for label-based monitoring metric calculation
- Agreement (Sec 2.1): For sample-based measurement definition, starting time = first detected path rounded down with timing granularity T for gNB/TRP measurement
- Conclusion (Sec 3.1): For measurement report of AI/ML assisted positioning Case 3a, LOS/NLOS indicator can't be reported independently from other measurements