UE AI Model Deployment via Capability Reporting
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
There is a need for an efficient method to train and deploy AI/ML models, particularly in scenarios where user equipment (UE) needs to indicate its capabilities for AI and ML, and then acquire and generate appropriate models for encoding and decoding processes.
Innovation Solution
A method where UE reports capability information to a network device, which then sends model information for an encoder and decoder model. The UE generates these models based on the received information and deploys the decoder model by sending its model information back to the network device for deployment.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If UE reports capability information and acquires models from network device, then model deployment efficiency is improved, but system complexity increases
Solution Approach 1:
The system divides model training and deployment into separate phases and components: capability reporting phase, model acquisition phase, model generation phase, and deployment phase. This segmentation allows each component to be optimized independently and improves overall system manageability despite increased complexity.
Solution Approach 2:
The network device provides model information in advance before the actual model generation and deployment. This preliminary provision of model information enables the UE to prepare and generate models efficiently, improving deployment speed while following a structured complexity management approach.
2Adaptability or versatility
If UE generates encoder and decoder models based on received model information, then encoding and decoding functionality is achieved, but processing time increases
Solution Approach 1:
The network device provides pre-processed model information that includes necessary parameters and structures. This preliminary preparation reduces the processing time the UE needs to spend on model generation, as the UE only needs to instantiate and configure the model based on the provided information rather than building it from scratch.
Solution Approach 2:
The UE generates models by copying and instantiating model structures and parameters provided by the network device. This copying approach allows the UE to create functional models quickly without reinventing the entire model architecture, thus achieving encoding and decoding functionality with reduced processing time.
3Productivity
If UE sends decoder model information back to network device for deployment, then model deployment is enabled, but communication overhead increases
Solution Approach 1:
The system extracts and transmits only the essential model information required for deployment rather than the entire model. This selective transmission approach enables model deployment capability while minimizing communication overhead by sending only necessary parameters and metadata.
Solution Approach 2:
The UE sends decoder model information back to the network device as feedback to enable deployment. This feedback mechanism allows the network device to receive and process the model information for deployment purposes, enabling a coordinated deployment process that balances communication requirements with deployment capabilities.
Data Source
AI summary
A model training and deploying method, performed by a user equipment (UE), includes: reporting capability information to a network device, the capability information being used for indicating support capabilities of the UE for at least one of an Artificial Intelligence (AI) or a Machine Learning (ML); acquiring at least one of model information of an encoder model to be trained or model information of a decoder model to be trained, sent by the network device; generating the encoder model and the decoder model based on at least one of the model information of the encoder model to be trained or the model information of the decoder model to be trained; and sending the model information of the decoder model to the network device, the model information of the decoder model being used for deploying the decoder model.