Radio Digital Twin Observability Engine for 5G Networks
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Monitoring and analyzing the operating conditions and performance parameters of large-scale 5G O-RAN cellular networks is computationally intensive and time-consuming due to the vast number of interconnected nodes and software/hardware modules, making it challenging to efficiently track and respond to anomalies in real-time.
Innovation Solution
A digital twin representation of the network is used in conjunction with trained machine learning models to identify a subset of nodes that reliably represent the entire network, allowing for efficient monitoring and anomaly detection with reduced computational resources and latency, enabling proactive identification and adaptation of operating conditions over time.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If monitoring and analyzing all nodes and modules in large-scale 5G O-RAN networks is performed, then complete network observability is achieved, but computational cost and time consumption increase significantly
Solution Approach 1:
The patent segments the large-scale radio network into multiple digital twin representations, each representing a subset of nodes and modules. The machine learning model processes these segmented digital twins independently and aggregates results, reducing the computational burden compared to processing the entire network as a single unit while maintaining comprehensive observability.
Solution Approach 2:
The patent extracts and identifies a representative subset of nodes and modules from the complete network using machine learning algorithms. This subset captures the essential operating conditions and performance parameters of the entire network, allowing monitoring efforts to focus on this smaller, representative group rather than every individual node.
2Measurement precision
If monitoring and analyzing all nodes and modules in large-scale 5G O-RAN networks is performed, then complete network observability is achieved, but time consumption increases significantly
Solution Approach 1:
The patent segments the monitoring task into parallel processing of multiple digital twin representations. Each digital twin can be processed independently and simultaneously, significantly reducing the total time required compared to sequential analysis of all network components.
Solution Approach 2:
The patent extracts a representative subset of network elements that can indicate overall network conditions. By monitoring this subset rather than all elements, the system achieves timely observability with reduced processing time while maintaining the ability to detect network-wide issues.
3Productivity
If a subset of digital twin representation is identified using machine learning, then computational resources and latency are reduced, but complexity of the system increases
Solution Approach 1:
The patent creates simplified digital twin copies of the physical network nodes and modules. These digital twins are virtual representations that can be processed more efficiently than the actual network elements. The machine learning model operates on these copies to identify representative subsets, reducing computational complexity while maintaining monitoring effectiveness.
Data Source
AI summary
A method for generating a digital twin of a radio network (RN) is disclosed that includes accessing, by one or more processing devices, a digital twin representation (DTR) of the RN, the RN including a plurality of nodes configured to enable communications between wireless user equipments (UEs), the DTR including virtual representations of the plurality of nodes and the interconnections. Using a trained machine learning (ML) model, a subset of the DTR is identified, the subset being representative of operating conditions for a particular time range. The method includes determining, by the processing devices via monitoring parameters of the subset of the DTR, that at least one of the operating conditions of the RN is outside of an expected range. In response to determining that the operating conditions of the RN is outside of the expected range, generating an output signal is generated indicative of a condition of the RN.