Wireless Positioning Signal Data Collection for Model Training
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing wireless communication systems, particularly 5G NR, lack the capability to support signaling for collecting training data to train positioning models based on measurements of positioning signals, and wireless devices are not configured to provide labeling assistance for such models.
Innovation Solution
Configuring wireless devices to collect and output training data for downlink positioning signals, including enhanced measurements and labels, to train a positioning model, and enabling network entities to transmit schedules and assistance information for improved positioning accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If wireless communication systems use existing signaling protocols, then system compatibility is maintained, but positioning model training capability is not supported
Solution Approach 1:
The patent segments the positioning functionality by separating model training data collection from standard positioning operations. It introduces dedicated information elements (IEs) specifically for model training (e.g., assistance information for model training, measurement reports for model training) that are distinct from regular positioning IEs. This allows the system to add positioning model training capability without modifying the core existing signaling protocols, thus maintaining compatibility while enabling new functionality.
Solution Approach 2:
The patent extracts the model training functionality from the standard positioning protocol framework. It creates separate, optional information elements that can be included or excluded based on whether model training is needed. This extraction approach allows systems that don't require model training to operate with standard protocols, while systems that do need model training can incorporate the additional elements without affecting other devices.
2Measurement precision
If wireless devices collect detailed positioning measurements and labels, then positioning model accuracy is improved, but data processing overhead increases
Solution Approach 1:
The patent implements preliminary action by having wireless devices pre-process positioning measurements and organize them into structured formats before transmission to the network. The devices perform initial filtering, formatting, and labeling of measurement data according to predefined templates. This preliminary processing reduces the complexity of subsequent network-side processing and ensures that the data is ready for immediate use in model training, thereby improving efficiency without sacrificing measurement precision.
Solution Approach 2:
The patent introduces an intermediary layer in the form of standardized information elements that mediate between raw measurements and the positioning model training process. These IEs serve as structured containers that organize measurement data, labels, and metadata in a standardized format. This intermediary structure simplifies data handling by providing a consistent interface between the wireless device, the network, and the machine learning model, reducing processing complexity while preserving measurement accuracy.
Data Source
AI summary
A wireless device may receive a set of positioning reference signals (PRSs). The wireless device may measure the set of PRSs. The wireless device may output at least one of a set of measurements based on the measured set of PRSs, a set of labels associated with the measured set of PRSs, or assistance information associated with the measured set of PRSs for training a positioning model. In one aspect, the wireless device may transmit, to a network entity, the set of measurements for the positioning model. In another aspect, the wireless device may include the positioning model. The wireless device may train the positioning model based on at least one of the set of measurements, the set of labels, or the assistance information. The wireless device may include a user equipment (UE) or a positioning reference unit (PRU). The network entity may include a location management function (LMF).