ML-Based Synchronization Anomaly Detection in Heterogeneous RAN Nodes

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Current synchronization troubleshooting tools in wireless communication networks struggle to detect, localize, and identify the root cause of synchronization issues among heterogeneous RAN nodes, often requiring manual intervention by domain experts.

Innovation Solution

A method and device utilizing Machine Learning models for anomaly detection and root cause analysis, which involves creating datasets from synchronized node information, running ML models for anomaly identification, and determining relationships between anomalous nodes and the network synchronization topology.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If manual analysis of alarm and performance logs is used to identify synchronization faults, then domain experts can identify probable causes, but the process is time-consuming and requires expert intervention

Engineering Contradiction:

Improvefault identification accuracyVSAvoidtroubleshooting time

Core Design Contradiction:

Measurement precisionVSLoss of time

Solution Approach 1:

The system performs self-diagnosis by automatically analyzing alarm and performance logs to identify synchronization faults and their root causes, eliminating the need for manual expert analysis. The automated fault identification system processes logs and generates diagnoses independently, allowing the network to troubleshoot itself without external intervention.

Inventive Principle:

Principle #25Self-service

Solution Approach 2:

The patent replaces manual mechanical analysis by domain experts with an automated electronic system that uses algorithms to process alarm and performance logs. This substitution transforms the troubleshooting process from human-centric manual analysis to machine-driven automated diagnosis, significantly reducing time while maintaining or improving accuracy.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

2Reliability

If GPS receivers are deployed at each site to ensure timing reliability, then timing synchronization is improved, but the cost increases significantly

Engineering Contradiction:

Improvetiming source reliabilityVSAvoiddeployment cost

Core Design Contradiction:

ReliabilityVSQuantity of substance

Solution Approach 1:

The patent enables existing RAN nodes to perform multiple functions, including synchronization fault detection and root cause analysis, without requiring dedicated GPS receivers at each site. The system uses centralized log analysis and automated diagnosis capabilities that can be deployed across the network infrastructure already in place, providing GPS-level reliability through software-based solutions rather than hardware additions.

Inventive Principle:

Principle #6Universality (Multi-functionality)

Solution Approach 2:

Instead of deploying physical GPS receivers at each RAN node, the system creates a virtual synchronization monitoring capability through automated analysis of performance logs and alarm data. The digital copy of synchronization status is obtained through log analysis, eliminating the need for physical GPS hardware while maintaining synchronization reliability.

Inventive Principle:

Principle #26Copying

3Adaptability or versatility

If RAN nodes use different technologies and compliance levels to support network heterogeneity, then adaptability is improved, but timing errors occur due to variations in standard compliance

Engineering Contradiction:

Improvenetwork heterogeneity supportVSAvoidsynchronization accuracy

Core Design Contradiction:

Adaptability or versatilityVSReliability

Solution Approach 1:

The patent introduces an intermediate automated diagnosis system that mediates between heterogeneous RAN nodes with different technologies and compliance levels. This intermediary system analyzes performance logs and alarm data to identify synchronization deviations caused by non-compliant nodes, enabling the detection and localization of timing errors without requiring all nodes to use identical technologies or compliance levels.

Inventive Principle:

Principle #24Intermediary (Mediator)

4Reliability

If synchronization paths are monitored for alarms, then active alarms can be detected, but synchronization misalignment due to delay asymmetries remains undetected

Engineering Contradiction:

Improvesynchronization path monitoringVSAvoidsynchronization misalignment detection

Core Design Contradiction:

ReliabilityVSDifficulty of detecting and measuring

Solution Approach 1:

The patent implements a feedback mechanism that continuously analyzes performance logs to detect synchronization misalignments caused by delay asymmetries. The system compares expected synchronization behavior with actual performance data, providing feedback that reveals hidden synchronization issues not detected by traditional alarm monitoring. This feedback loop enables the system to identify problems in synchronization paths even when no active alarms are present.

Inventive Principle:

Principle #23Feedback

Data Source

PatentEP4324124B1Device and method for synchronization among nodes

Publication Date: 2025.01.15 TELEFONAKTIEBOLAGET LM ERICSSON (PUBL)

Figure 1a~1b
Figure 2
Figure 3a

AI summary

The invention relates to a method and a device (101) for monitoring a synchronization status among a plurality of nodes (102-107) operating in a communications network, corresponding computer program (604) and computer program product (605). By using a Machine Learning, ML, model, datasets based on obtained (201) information associated with a plurality of nodes (102-107) and comprising information on the evolution of relationships between the plurality nodes (102-107) of the network, the device (101) may detect, localize, and identify anomalous nodes (102, 103, 107), time-slices when the anomalous nodes (102, 103, 107) exhibit anomalous behavior and a root of the anomalous behavior.