UE AI Prediction Model Control via Network Signaling
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In wireless communication systems, the efficient control and management of AI prediction models on User Equipment (UE) are hindered by the lack of clear starting and stopping criteria, leading to unnecessary power consumption and potential delays in obtaining prediction results, as existing solutions either consume excessive power or fail to provide timely predictions.
Innovation Solution
Implementing a method where control information is used to configure and control the operation of a prediction model on the UE, including threshold information for starting and stopping, allowing the UE to autonomously manage its AI prediction functions based on time, position, movement speed, signal quality, and prediction thresholds, thereby optimizing power usage and prediction timeliness.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the AI prediction model operates continuously on the UE, then prediction results are always available, but power consumption increases
Solution Approach 1:
The prediction model operates periodically rather than continuously. The network device sends control information at specific intervals or triggered by specific events (such as handover decisions), causing the UE to execute the prediction model only when needed. This periodic operation maintains prediction availability while significantly reducing power consumption compared to continuous operation.
Solution Approach 2:
The UE autonomously determines when to execute the prediction model based on control information from the network device. The control information contains configuration parameters that enable the UE to self-manage the prediction model operation without requiring constant network intervention, thereby reducing signaling overhead and power consumption while ensuring predictions are made at appropriate times.
2Reliability
If the AI prediction model operates continuously on the UE, then prediction results are always available, but device complexity increases
Solution Approach 1:
The UE is configured with control information containing all necessary parameters to autonomously manage the prediction model operation. The UE uses these parameters to determine when and how to execute predictions without requiring complex continuous network control or manual configuration, simplifying the overall system while ensuring prediction availability.
Solution Approach 2:
The network device pre-configures the UE with control information including configuration parameters before the UE needs to perform predictions. This preliminary configuration enables the UE to autonomously execute predictions at the right moments without requiring complex real-time decision-making or continuous network interaction, thereby reducing device complexity.
3Ease of operation
If the prediction model operates without control information, then operation is simple, but prediction timeliness deteriorates
Solution Approach 1:
The network device sends control information in advance containing configuration parameters that guide the UE's prediction model operation. This preliminary configuration enables the UE to execute predictions at the optimal moment without waiting for continuous commands, maintaining operational simplicity while ensuring timely predictions for handover decisions.
Solution Approach 2:
The control information from the network device provides feedback guidance to the UE on when and how to execute predictions. Based on this feedback, the UE can autonomously determine the appropriate timing for prediction execution, ensuring predictions are made timely without requiring complex continuous control or sacrificing operational simplicity.
Data Source
AI summary
A method for transmitting information is performed by a User Equipment (UE), and includes: receiving control information, wherein the control information is at least configured to control an operation of a first prediction model in the UE, and the first prediction model is configured to obtain a prediction result of Radio Resource Management (RRM).