5G Network Function Telemetry Zooming for Faster Fault Isolation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The challenges of flexible data collection for monitoring and telemetry applications in 5G networks, and the difficulty in detecting network issues remotely from their cause, lead to inefficiencies in debugging and optimization, particularly in managing the complexity and dynamism of 5G networks.
Innovation Solution
A machine-learning based network analyzer collects module-level statistics, identifies deviations from normal operation using a trained model, and provides suggestions for debugging or optimization by zooming in on specific network stages to address issues.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If dedicated processing hardware is deployed for each base station, then processing reliability is improved, but deployment cost and device complexity increase
Solution Approach 1:
The patent creates virtual copies of base station processing functions through virtualized network functions (VNFs) that run on generic hardware platforms. Instead of deploying physical dedicated hardware for each base station, the system instantiates software-based copies of processing functions (PHY, MAC, RLC, PDCP layers) that can be replicated and migrated across multiple physical hosts, thereby maintaining reliability through redundancy while eliminating the need for complex dedicated hardware deployments.
Solution Approach 2:
The patent replaces the mechanical/physical system of dedicated hardware base stations with a software-based virtualized system. The processing functions are transformed from hardware-bound operations to software workloads that can be dynamically allocated, managed, and scaled on generic computing infrastructure, thus reducing deployment complexity while maintaining processing capabilities.
2Adaptability or versatility
If virtualized RAN with generic computing resources is used, then scalability and fault tolerance are improved, but network issue detection capability deteriorates
Solution Approach 1:
The patent introduces a dedicated monitoring and management system that acts as an intermediary between the virtualized network functions and the operational support system. This intermediary collects performance metrics, logs, and operational data from multiple VNF instances across the virtualized RAN, aggregates this information, and presents it in a unified manner for analysis, thereby enabling effective network issue detection despite the distributed and virtualized nature of the infrastructure.
Solution Approach 2:
The patent implements continuous feedback loops where performance metrics from virtualized network functions are constantly monitored, analyzed, and fed back to the management system. This feedback mechanism enables real-time detection of anomalies, performance degradation, and faults in the virtualized environment, allowing operators to respond to network issues promptly despite the complexity of the virtualized architecture.
3Measurement precision
If comprehensive data collection is performed across all network functions, then measurement precision is improved, but data processing overhead and loss of time increase
Solution Approach 1:
The patent extracts and collects only the specific performance metrics and operational data that are most relevant for monitoring and troubleshooting virtualized network functions. Instead of gathering all possible data from every VNF, the system identifies key performance indicators (KPIs) such as throughput, latency, packet loss, and resource utilization that are critical for detecting network issues, thereby reducing data volume while maintaining monitoring precision.
Solution Approach 2:
The patent segments the data collection process by organizing metrics according to network layers (PHY, MAC, RLC, PDCP) and functional units. This segmentation allows the monitoring system to process and analyze data in manageable chunks, applying different analysis methods to different segments, thereby reducing overall processing time while maintaining comprehensive monitoring coverage across all network functions.
Data Source
AI summary
A system, method, and computer-readable media for analyzing network traffic in a communications network are provided. A network analyzer collects a subset of module-level statistics from a plurality of network functions (NFs) within the network. The network analyzer analyzes the module-level statistics to identify a deviation from a normal state of operation. The network analyzer identifies a network stage where a network issue causes the deviation using a machine learning model trained on features associated with the deviation. The network analyzer provides suggestions for debugging, optimizing, or testing the network issue based on the deviation and the identified network stage. The network analyzer can collect additional module-level statistics or packet-level statistics at the identified network stage based on the deviation, confirm the network issue based on the additional module-level statistics or packet-level statistics, and take corrective action to an NF at the network stage to address the network issue.