SD-WAN Path Anomaly Detection via Time-Series Dynamics
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In large-scale enterprise software-defined wide area networks (SD-WANs), it is challenging for network administrators to identify and rectify anomalous paths that do not provide a satisfactory user experience for online/software-as-a-service (SaaS) applications due to the complexity of tens of thousands of paths and varying network dynamics.
Innovation Solution
A device computes time-series dynamics for network performance metrics, categorizes them, and determines if paths are anomalous, providing an indication for display, allowing for automatic detection and tracking of experience-degrading paths using KPI dynamics analysis and predictive routing processes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If manual path identification and rectification methods are used in large-scale SD-WANs, then network administrators can identify and fix anomalous paths, but the complexity of managing tens of thousands of paths makes this process extremely challenging and time-consuming
Solution Approach 1:
The system enables automatic self-diagnosis and self-identification of anomalous paths through machine learning models that autonomously analyze network performance data, eliminating the need for manual intervention in path identification and classification
Solution Approach 2:
Manual mechanical analysis of network paths is replaced with automated computational systems using time-series dynamics analysis and machine learning algorithms to detect, classify, and track anomalous paths at scale
2Ease of repair
If network administrators manually interact with multiple network and service providers to rectify path issues, then problems can be resolved, but this process is extremely challenging and slows down rectification
Solution Approach 1:
The system performs preliminary classification and identification of anomalous paths before rectification is needed, pre-processing network data to categorize paths by anomaly type, which accelerates the subsequent rectification process by providing organized information ready for action
Solution Approach 2:
The patent introduces an automated intermediary system that acts as a mediator between network administrators and multiple service providers, streamlining communication and coordination for path rectification through centralized anomaly management
3Extent of automation
If traditional network monitoring methods are used, then basic network status can be monitored, but automatic detection and tracking of experience-degrading paths with varying dynamics is not achieved
Solution Approach 1:
The system transforms network performance monitoring by changing from static threshold-based detection to dynamic time-series analysis, using multiple performance parameters (latency, jitter, packet loss) analyzed through machine learning to detect anomalies that vary over time
Solution Approach 2:
The patent applies dynamic analysis methods that adapt to varying network conditions, using time-series dynamics and machine learning models that continuously learn from changing network patterns to automatically detect and track anomalous paths with varying characteristics
Data Source
AI summary
In one embodiment, a device computes time series dynamics for a performance metric of a path in a network used to convey traffic for an online application. The device matches those time series dynamics to one or more dynamics categories. The device makes a determination as to whether the path in the network is anomalous, based on the one or more dynamics categories. The device provides, based on the determination, an indication that the path in the network is anomalous for display.