UE Capability Signaling for DL-RFFP Positioning Accuracy

The patent implements dynamic capability negotiation between UE and location server through provide capabilities messages. The UE dynamically indicates its DL-RFFP capabilities (machine learning model support, neural network support, fingerprint database support) and the location server dynamically provides appropriate assistance data based on these capabilities, allowing the system to adapt to different UE configurations and optimize positioning performance.

The patent changes the parameter state of UE capabilities by introducing specific capability indicators for DL-RFFP functions. The UE transitions from a generic capability state to a specific DL-RFFP capability state by providing detailed capability information (support for machine learning models, neural networks, fingerprint databases), enabling the system to leverage these specific parameters for enhanced positioning accuracy.

The patent applies preliminary action by having the location server prepare and provide assistance data (including machine learning models and neural networks) to the UE in advance of the actual positioning measurement. The UE receives and stores these models beforehand, so when positioning is needed, the pre-prepared models can be immediately applied to fingerprint matching, reducing real-time processing complexity while maintaining high accuracy.

The patent introduces machine learning models and neural networks as intermediaries between the raw RF signal measurements and the positioning calculation. These intermediary models handle the complex pattern recognition and fingerprint matching tasks, allowing the UE to achieve high positioning accuracy without directly implementing complex algorithms, thereby reducing overall device complexity.