AI/ML Model Lifecycle Management in 5G Wireless Networks
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current 5G wireless communication networks lack effective life cycle management for AI/ML models, which are deployed in terminals, leading to inefficiencies in resource utilization and performance optimization.
Innovation Solution
The proposed solution involves methods and systems where user equipment (UE) reports its AI/ML capabilities to a base station, allowing the network to configure and manage AI/ML operations, including measurement and reporting configurations, enabling the network to assist in life cycle management of AI/ML models.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Extent of automation
If AI/ML models are deployed in terminals without network assistance, then device autonomy is improved, but resource utilization efficiency deteriorates
Solution Approach 1:
The network acts as an intermediary between AI/ML model developers and terminals. The network receives capability information from terminals, determines appropriate AI/ML models, and provides them to terminals. This mediator approach allows terminals to operate autonomously while the network optimizes resource utilization by selecting and distributing appropriate models.
Solution Approach 2:
The system dynamically manages AI/ML model deployment based on terminal capabilities and network conditions. Terminals can report their AI/ML capabilities to the network, and the network can provide configuration information and update models as needed. This dynamic approach balances device autonomy with network-controlled resource optimization.
2Productivity
If network assists in life cycle management of AI/ML models, then resource allocation is improved, but system complexity increases
Solution Approach 1:
The life cycle management of AI/ML models is segmented into distinct phases: capability reporting by terminals, model selection by the network, configuration provision, and model execution at terminals. This segmentation distributes complexity across different network entities and phases, making the overall system more manageable despite the increased coordination required.
Solution Approach 2:
The system implements feedback loops where terminals report their AI/ML capabilities to the network, and the network provides configuration information and model updates based on these reports. This feedback mechanism enables the network to optimize resource allocation while managing complexity through structured information exchange and standardized procedures.
Data Source
AI summary
The disclosure relates to a 5th generation (5G) or 6th generation (6G) communication system for supporting a higher data transmission rate. A method performed by a user equipment (UE) in a communication system is provided. The method includes transmitting, by the UE to a base station, capability information indicating a set of artificial intelligence (AI)/machine learning (ML) functionalities, receiving, by the UE from the base station, configuration information associated with an AI/ML inference, wherein the configuration information indicates at least one of a measurement configuration or a reporting configuration, receiving, by the UE from the base station, information to indicate activation of an AI/ML functionality, and performing, by the UE, an AI/ML based operation based on the configuration information.