O-RAN High Availability Management With Dynamic Redundancy
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current approaches to achieving high availability in Telco RAN involve adding redundant resources, which are not efficiently utilized, leading to increased costs in Telco cloud deployment.
Innovation Solution
Implementing a hierarchical high availability management system using local and centralized managers, coupled with AI/ML algorithms, to monitor and dynamically manage redundant resources across RF, computation, and interface clusters in O-RU and O-DU, enabling efficient load balancing and failure prediction.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If redundant resources are added to achieve high availability in Telco RAN, then system reliability is improved, but deployment cost increases due to inefficient resource utilization
Solution Approach 1:
The patent implements a hierarchical high availability manager with AI/ML-based failure prediction that continuously monitors resource states and provides feedback for dynamic resource allocation. This feedback mechanism enables the system to optimize resource utilization by allocating redundant resources only when and where failures are predicted, rather than maintaining static redundancy across all resources.
Solution Approach 2:
The system dynamically adjusts resource allocation based on real-time conditions and AI/ML predictions. The high availability manager can flexibly provision or de-provision redundant resources according to changing demands and predicted failure risks, transforming the static redundancy model into a dynamic one that adapts to actual system needs.
2Reliability
If redundant resources are added to achieve high availability in Telco RAN, then system reliability is improved, but deployment cost increases
Solution Approach 1:
The AI/ML-based failure prediction capability performs preliminary identification of resources at risk before actual failures occur. This allows the system to proactively allocate redundant resources only to specific components predicted to fail, rather than maintaining universal redundancy, thereby reducing overall deployment costs while maintaining high availability.
Solution Approach 2:
The system changes the parameter of resource allocation from a static, uniform approach to a dynamic, differentiated approach based on AI/ML analysis. By changing how resources are allocated (from fixed redundancy to adaptive provisioning), the system achieves high availability with reduced total resource requirements and lower deployment costs.
3Loss of energy
If hierarchical high availability management with AI/ML is implemented, then resource utilization efficiency is improved, but system complexity increases
Solution Approach 1:
The patent divides the high availability management system into hierarchical levels with specialized functions. The AI/ML-based failure prediction and dynamic resource allocation are implemented as separate modular components that can be independently developed, deployed, and maintained. This segmentation reduces overall system complexity by breaking down the complex management task into manageable, specialized modules.
Solution Approach 2:
The hierarchical high availability manager acts as an intermediary layer between the physical resources and the control plane. This intermediary component abstracts the complexity of AI/ML-based resource management, providing a standardized interface for resource allocation while handling the complex prediction and optimization logic internally, thereby shielding the rest of the system from complexity.
Data Source
AI summary
System and method embodiments are disclosed for high availability management for open radio access network (O-RAN). The O-RAN may be deployed on cloud with the O-CU deployed on a region cloud, O-RUs deployed on a cell site O-Cloud, and O-DUs deployed on an edge cloud. Each O-RU may comprise one or more RF clusters, computation clusters, and interface clusters. O-RU instances and O-DU instances may be instantiated with redundancy on the cell site O-Cloud and on the edge cloud, respectively, to serve one or more users. Local and central high-availability (HA) managers may be used to monitor O-RU instance performance for failure prediction/detection and to monitor internal states of each O-DU instance. In response to O-RU instance failure or O-DU internal states beyond/below state thresholds, new O-RU or O-DU instances may be instantiated as replacement instances for O-Cloud high availability management.