RAN Intelligent Controller Segmentation for Latency Reliability
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Next-generation wireless networks face challenges in optimizing resource allocation and performance metrics such as latency and reliability across different time scales, particularly in Centralized Radio Access Networks (C-RAN) and Open Radio Access Networks (O-RAN) architectures, where existing solutions struggle to balance latency and reliability effectively.
Innovation Solution
The implementation of a RAN Intelligent Controller (RIC) system that includes non-real-time, near-real-time, and real-time RIC components, each utilizing AI/ML models to manage radio resources based on specific time scales and performance metrics, optimizing network slicing, radio resource management, and application service support through open interfaces and machine learning models.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If a RAN Intelligent Controller (RIC) system with multiple time-scale components is implemented, then resource allocation optimization and network capacity are improved, but system complexity increases
Solution Approach 1:
The RIC system is segmented into three distinct time-scale components: non-real-time RIC (strategic planning), near-real-time RIC (tactical control), and real-time RIC (operational execution). Each component handles specific functions appropriate to its time scale, enabling optimized resource allocation while managing complexity through functional decomposition rather than monolithic design
Solution Approach 2:
The system introduces a temporal dimension to resource management by organizing control functions across multiple time scales. This dimensional approach allows simultaneous optimization of long-term planning, medium-term coordination, and short-term real-time adjustments without creating a single complex control loop
2Adaptability or versatility
If AI/ML models are utilized for radio resource management, then adaptability to different time scales is improved, but computational requirements and system complexity increase
Solution Approach 1:
Different AI/ML models are deployed at different time scales with appropriate local characteristics: non-real-time RIC uses models for strategic planning with longer training periods, near-real-time RIC employs models for tactical adjustments, and real-time RIC utilizes models for immediate resource allocation. Each layer's model is optimized for its specific temporal requirements rather than using a single universal model
Solution Approach 2:
The RIC architecture serves as an intermediary layer between network resources and application services, coordinating AI/ML computations across different time scales. This intermediary structure enables adaptability to various time requirements while distributing computational burden across multiple components rather than concentrating it in a single system
3Productivity
If network slicing and radio resource management are optimized, then user throughput and connectivity are improved, but control mechanism complexity increases
Solution Approach 1:
The RIC system dynamically adjusts resource allocation across network slices based on real-time conditions and time-scale requirements. The system can dynamically create, modify, and terminate network slices, and dynamically adjust radio resources within slices based on current network state and service requirements, enabling optimized throughput without static complex configurations
Data Source
AI summary
A method, a device, and a non-transitory storage medium are described in which a radio service is provided. The radio service may provide radio resource control of radio resources of at least three different time scales. The radio service may include machine learning or artificial intelligent devices that include radio network information of the at least three different time scales.