R1-2409841
discussion
Discussion on specification support for positioning accuracy enhancement
Summary
Ruijie Networks presents 4 proposals for AI/ML positioning accuracy enhancement in NR air interface, focusing on sample-based measurements, model output reporting, training data collection, and model monitoring for various AI/ML positioning cases.
Position
Ruijie Networks advocates for maximizing reuse of existing information elements and frameworks rather than creating new ones, supporting flexible options for sample-based measurements that preserve potentially useful information before first detected path, and establishing mandatory timing/timestamp reporting while keeping quality indicators optional for implementation flexibility.
Key proposals
- Proposal 1 (Sample-based vs path-based measurements): For sample-based channel response measurements between UE and TRP, further down-select between Option A (starting time based on first detected path timing) and Option B (starting time based on earliest sample where first path power is detectable)
- Proposal 2 (Model output for assisted positioning): For AI/ML positioning Case 3a timing reports from gNB to LMF, mandate timing information and timestamp, make optional the quality of timing information and LOS/NLOS indicator, reusing existing IEs
- Proposal 3 (Training data collection): For AI/ML positioning Cases 1, 3a and 3b label reporting, reuse existing IE LocationUncertainty-r16 from TS 37.355 for quality indicator representation
- Proposal 4 (Model monitoring): For AI/ML positioning Case 1 label-based model monitoring, support Option A-3 as baseline using Rel-18 assistance data transfer framework from LMF to target UE with PRU measurements and locations