R1-2410043
discussion
On AI/ML-based CSI compression
From InterDigital
Summary
InterDigital presents comprehensive analysis of AI/ML-based CSI compression for NR air interface, covering beam domain processing, temporal-spatial-frequency compression, model monitoring, and inter-vendor collaboration. The document contains 25 observations and 5 proposals addressing performance improvements, complexity reduction, and specification impacts for Release 19 standardization.
Position
InterDigital advocates FOR beam domain processing as a superior approach to spatial domain processing, offering better generalization across antenna configurations and reduced complexity. They strongly push FOR temporal-spatial-frequency (TSF) compression Case 2 over spatial-frequency only compression, demonstrating up to 33% overhead reduction and significant cell-edge throughput gains (12-17% over SF, 25-45% over Rel-16). They promote comprehensive UE-side model monitoring with multiple trigger mechanisms and metrics beyond SGCS, while supporting Option 3a-1 for inter-vendor collaboration as a feasible solution with acceptable performance degradation.
Key proposals
- Proposal 1 (Sec 2.2): Study the specification impact of adapting the length of the historical CSI buffer to channel conditions, such as UE speed
- Proposal 2 (Sec 2.4): Study further the following aspects for model monitoring in Rel-19: Details of reporting mechanism for the monitoring metrics with both time/event-trigger based, Appropriate UE-side monitoring metric which reflects AI/ML model performance accurately, UE-side monitoring based on precoded RS (CSI-RS, DM-RS), Reporting contents/structure of UE-side monitoring metric and its associated feedback overhead, NW-side monitoring with lower signaling overhead
- Proposal 3 (Sec 2.4): Mechanisms for identifying the cause of performance degradation include: UE assistance information such as UE-side model monitoring metrics and UE-side out-of-distribution metrics, Reporting the UE-assessed error cause (e.g., data drift, UE-side, or undetermined), Mitigation mechanisms, including fallback to legacy CSI reporting, model switching
- Proposal 4 (Sec 3.4.2.1): TSF compression performance should be evaluated under multiple observation window lengths
- Proposal 5 (Sec 3.4.2.3): For further performance improvement of AI-TSF compression Case2 with missing UCI, study different missing UCI mitigation solutions e.g., partial buffer reset, or retransmission of the missed UCI