UE RRM Prediction Model for Wireless Load Reduction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In wireless communication systems, existing technologies face challenges in providing accurate Radio Resource Management (RRM) predictions on user equipment (UE) due to the need for network-side data storage and calculation, which compromises user experience and data security, and increases wireless communication load.
Innovation Solution
Implementing a prediction model on the UE side that uses local historical data to determine RRM prediction results, eliminating the need for network-side data storage and calculation, and allowing for customized AI module training based on local data, thereby improving prediction accuracy and security while reducing communication load.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Extent of automation
If network-side data storage and calculation are used for RRM predictions, then centralized control is achieved, but user experience and data security deteriorate and wireless communication load increases
Solution Approach 1:
The patent inverts the traditional architecture by moving the prediction model from the network side to the user equipment (UE) side. Instead of the network storing data and making predictions centrally, the UE locally stores historical data and executes the prediction model, thereby improving data security and reducing communication load while maintaining prediction functionality.
Solution Approach 2:
The UE performs RRM predictions autonomously using its own historical data and the prediction model. The network only needs to send the model parameters, and the UE independently generates prediction results, reducing the need for continuous network-side processing and data transmission.
2Extent of automation
If network-side data storage and calculation are used for RRM predictions, then centralized control is achieved, but wireless communication load increases
Solution Approach 1:
The patent extracts the data storage and calculation functions from the network side and places them on the UE side. The network only transmits the prediction model parameters, while the UE handles historical data storage and prediction calculations locally, significantly reducing wireless communication load.
Solution Approach 2:
Instead of transmitting actual historical data from UE to network for processing, the patent copies the prediction model to the UE and allows it to generate prediction results locally, avoiding the need to transmit large amounts of historical data over the wireless channel.
3Reliability
If prediction model is implemented on UE side using local historical data, then data security and prediction accuracy improve, but device complexity increases
Solution Approach 1:
The patent transforms the complex task of handling large historical datasets into a simpler parameter-based approach. The network provides pre-processed model parameters, and the UE only needs to execute the prediction model with these parameters and local historical data, reducing the computational burden on the UE.
4Adaptability or versatility
If prediction model is implemented on UE side using local historical data, then customized AI module training is enabled, but device complexity increases
Solution Approach 1:
The network performs preliminary processing by training the prediction model offline and extracting model parameters before sending them to the UE. This preliminary action reduces the complexity of on-device implementation, as the UE only needs to execute the pre-trained model with local data rather than performing full model training.
Data Source
AI summary
An information transmission method is performed by a user equipment (UE), and includes: determining, through a prediction model run by the UE, a prediction result of radio resource management (RRM) according to configuration information; and reporting the prediction result to an access network device according to the configuration information.