Non-RT RIC AI/ML Training Service API Design
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The Open Radio Access Network (O-RAN) lacks detailed specifications for Artificial Intelligence (AI)/Machine Learning (ML) training services in Non-RT RIC, specifically in terms of use cases, procedures, and RESTful API designs, which are essential for integrating AI/ML capabilities into wireless communication networks.
Innovation Solution
The patent provides detailed techniques and designs for AI/ML training services in Non-RT RIC, including service operations like requesting training, querying training status, canceling training, and notifying training status, using RESTful API designs that conform to the principles of the Representational State Transfer (REST) architectural style, and defining resource URI structures and methods for training jobs and status notifications.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If detailed specifications for AI/ML training services are added to Non-RT RIC, then the capability to provide policy-based guidance and enrichment for intelligent RAN optimization is improved, but the device complexity increases
Solution Approach 1:
The patent segments the AI/ML training service into distinct operational components: service request reception, training job creation, status querying, and notification mechanisms. Each component is handled through separate RESTful API endpoints, allowing the Non-RT RIC to manage complex AI/ML training services through modular, manageable operations that reduce overall system complexity while maintaining full functionality
Solution Approach 2:
The patent introduces RESTful API interfaces as intermediary layers between the Non-RT RIC and AI/ML training services. These standardized interfaces act as mediators that simplify the interaction complexity by providing uniform methods (POST, GET, DELETE) for service management, thereby enabling advanced AI/ML capabilities without proportionally increasing operational complexity
2Ease of operation
If RESTful API designs and procedures are implemented for AI/ML training services, then the ease of operation is improved, but the device complexity increases
Solution Approach 1:
The patent implements universal RESTful API designs that can handle multiple AI/ML training service operations (request creation, status querying, cancellation, notifications) through standardized HTTP methods and message formats. This universal approach allows the Non-RT RIC to manage diverse AI/ML training services using consistent procedures, improving ease of operation while the standardized nature of RESTful APIs prevents excessive complexity increase
Data Source
AI summary
A machine-readable storage medium, an apparatus and a method, each corresponding to either a service consumer or a service producer of a non-real-time (non-RT) radio access network intelligent controller (RIC) of a Service Management and Orchestration Framework (SMO FW). Communications from the service consumer to the service producer include: a training request for artificial intelligence/machine learning (AI/ML) training job; a query regarding a training status of the AI/ML training job; a cancel training request to cancel the AI/ML training job; and a notification regarding the training status of the AI/ML training job.