Digital Network Assistant for Anomaly Remediation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Managing complex networks, such as telecommunications and IoT networks, is challenging due to the impracticality of individual engineers diagnosing and remediating network faults across large scales, which often involve multiple possible causes like misconfiguration or equipment failure.
Innovation Solution
A digital network assistant that ingests data streams from network elements, analyzes key performance indicators, identifies anomalies, and recommends actions to remediate them, either automatically or through an augmented reality interface allowing users to perform physical changes on network elements.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If individual engineers are used to diagnose and remediate network faults, then complete knowledge of the network can be achieved, but it becomes impractical as network scale increases
Solution Approach 1:
The system enables automated self-diagnosis and self-remediation of network faults through machine learning models that autonomously analyze network data, identify anomalies, and execute remediation actions without requiring human engineer intervention for each incident
Solution Approach 2:
An automated system acts as an intermediary between network monitoring and fault remediation, using trained machine learning models to process network data, identify faults, and coordinate remediation actions, thereby bridging the gap between comprehensive monitoring and practical operational feasibility
2Adaptability or versatility
If the number of engineers increases to manage larger networks, then more network knowledge is available, but it becomes impractical for individual engineers to diagnose faults across the entire network
Solution Approach 1:
The machine learning-based diagnostic system serves multiple functions including anomaly detection, fault classification, root cause analysis, and remediation recommendation, replacing the need for multiple specialized engineers while maintaining comprehensive network management capability
Solution Approach 2:
The patent replaces the mechanical system of human engineers manually diagnosing faults with an automated computational system that uses machine learning algorithms to analyze network data and identify faults, thereby reducing operational complexity while maintaining or improving diagnostic capability
3Productivity
If automated actions are performed for high-confidence remediation, then response time is reduced, but manual review may be needed for complex anomalies
Solution Approach 1:
The system dynamically adjusts the confidence threshold parameter based on anomaly characteristics, allowing fully automated remediation for high-confidence cases while triggering manual review for lower-confidence or complex anomalies, thereby optimizing both response time and remediation accuracy
Data Source
AI summary
A digital network assistant which can detect network anomalies, identify actions likely to remediate them, and assist the user in carrying out those actions. In particular, a digital network assistant constantly monitors data streams associated with the network to determine key performance indicators for the network. When these key performance indicators indicate a network anomaly, the digital network assistant associates it with a digital string to one or more actions likely to remediate similar network issues. The digital network assistant can take these actions automatically or present them to a user to be taken. The system can also aid the user in taking the required actions via an augmented reality interface. In addition, the system can create narratives embedding findings from data analysis eliminating subjectivity. The system can also find optimal parameter sets by continuously analyzing anomaly-free parts of the network and their key performance indicators.