R1-2410151
discussion
AI/ML based CSI Prediction
From Google
Summary
Google proposes enhancements for AI/ML based CSI prediction in NR air interface, focusing on UE-side model consistency between training and inference. The document contains 2 main proposals addressing CSI-RS configuration reporting and associated ID management for CSI report configurations.
Position
Google advocates FOR UE-side only CSI prediction models with UE-reported preferred CSI-RS configurations to avoid configuration mismatches, and FOR associated ID management per CSI report configuration to maintain training/inference consistency. They push FOR dynamic activation/deactivation capabilities to enable network energy saving while maintaining model performance.
Key proposals
- Proposal 1: Support the UE reports the preferred CSI-RS configuration for CSI prediction including at least the preferred intervals between every two consecutive CSI-RS instances and minimum number of CSI-RS instances for CSI prediction
- Proposal 2: Support to configure associated ID per CSI report configuration and support dynamic activation/deactivation of a CSI report configuration