AI Network Function Data Allocation for Unified Service Delivery
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing AI service provision methods lack a unified technical framework, leading to fragmentation of network AI application scenarios and inefficiencies due to the lack of personalized data management and algorithm categorization, resulting in low efficiency and poor convenience.
Innovation Solution
A method involving a first AI network function (NF) with decision-making capability, which receives a request message, determines data allocation, and collaboratively executes AI services with multiple AI NF elements, forming a unified framework for AI service provision.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If AI services are provided based on optimization and transformation on traditional network architecture, then AI technology can be integrated into wireless communication networks, but the network AI application scenarios become fragmented and lack a unified technical framework
Solution Approach 1:
The patent segments the AI service provision process into distinct functional modules: AI service request reception module, data type determination module, AI algorithm selection module, and AI service execution module. This segmentation allows each module to handle specific tasks independently while maintaining overall system unity through standardized interfaces and a unified technical framework that coordinates these segments.
2Ease of manufacture
If AI service provision processes simply superimpose on existing network processes, then implementation is straightforward, but personalized AI service needs cannot be met and efficiency is low
Solution Approach 1:
The patent implements dynamic AI service provision where the system adaptively determines data types (structured or unstructured) based on service requirements, dynamically selects appropriate AI algorithms from a repository, and flexibly allocates computational resources. This dynamic approach allows the system to optimize for personalized service needs while maintaining operational efficiency, moving beyond static superposition of AI on traditional networks.
3Device complexity
If data types and AI algorithms are not subdivided and categorized, then the system structure remains simple, but management efficiency and convenience deteriorate
Solution Approach 1:
The patent applies parameter changes by categorizing data into distinct types (structured data with predefined schemas and unstructured data without fixed schemas) and organizing AI algorithms by their functional parameters and适用 scenarios. This parameter-based classification system enables efficient management and selection of data and algorithms without requiring complex ad-hoc analysis, improving operational convenience while maintaining reasonable system structure.
Data Source
AI summary
A method for providing a service based on AI is executed by a first AI network function (NF) network element with decision-making capability. The method includes: receiving a first request message sent by an AMF network element indicating an AI service requested by a terminal device; determining a data allocation result based on the first request message, in which the data allocation result includes: indication information corresponding to data allocated to a first AI NF network element and at least one second AI NF network element; sending the data allocation result to the NF network element, and receiving data allocated by the NF network element based on the data allocation result, and determining a first data execution result based on the data allocated by the NF network element, and receiving a second data execution result sent by the at least one second AI NF network element.