R1-2409854
discussion
Discussion on CSI compression
From NEC
Summary
NEC presents their views on AI/ML-based CSI compression for NR air interface, focusing on inter-vendor training collaboration solutions and specification impacts. The document contains 10 proposals and 3 observations addressing training collaboration methods, performance monitoring approaches, and model inference procedures.
Position
NEC strongly advocates FOR Direction A (parameter/model/dataset sharing enabling UE-side offline engineering) as the priority approach for inter-vendor collaboration, arguing it provides better performance and reduces NW complexity compared to other directions. They push AGAINST Direction C (fully standardized reference models) citing performance limitations and data distribution misalignment issues, and express concerns about Direction B's complexity in managing UE-specific models at the network side.
Key proposals
- Proposal 1 (Sec 2): RAN1 to prioritize studying Direction A: Sharing parameters/reference model/dataset that enables UE-side offline engineering (Inter vendor collaboration option 3a/5a/4)
- Proposal 2 (Sec 3): For Option 3a/4-1, discuss corresponding Pros/Cons prior to make a choice between Alt.1 (UE-side first trains a nominal decoder) and Alt.2 (UE-side directly develops and trains the actual encoder)
- Proposal 3 (Sec 4): Support NW-side monitoring based on the target CSI reported by the UE via legacy eT2 codebook or eT2-like high-resolution codebook
- Proposal 4 (Sec 4): For NW-side monitoring, the AI CSI and associated target CSI can be reported in the same reporting instance, or two separate reports
- Proposal 5 (Sec 4): For NW-side monitoring, RAN1 to discuss and evaluate whether ground truth reporting from UE can be transmitted over L1 or higher layer signaling (e.g. RRC)
- Proposal 6 (Sec 4): Support UE-side monitoring based on the output of the CSI reconstruction model at the UE or via direct estimation of intermediate KPI without reconstructing a target CSI
- Proposal 7 (Sec 5): At least for Case 2, CSI buffer reset should be supported to address misalignment of historical CSI used at UE side and NW side
- Proposal 8 (Sec 5): At least for Case 2, further study effective availability of historical CSI information over time
- Proposal 9 (Sec 5): If the CSI reconstruction part/model at UE side is available, support Option 2a for CQI determination
- Proposal 10 (Sec 5): For defining the pairing information used to enable the UE to select a CSI generation model(s) that is compatible with the CSI reconstruction model(s) used by the gNB, down select from three model ID-based options