SD-WAN Application Flow Routing with Predictive SLA Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current SD-WAN networks experience significant delays in detecting Service Level Agreement (SLA) violations, leading to poor user experience, and the addition of new applications without proper planning can introduce unforeseen issues due to assumptions about usage and impact on existing applications.
Innovation Solution
A system that monitors application flows, uses machine learning to predict potential SLA violations, and proactively routes application flows to optimal paths, minimizing SLA violations by intelligently distributing flows across multiple paths.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If current SD-WAN networks use traditional SLA violation detection methods, then the network can detect violations, but the detection time is significant (minutes), leading to poor user experience
Solution Approach 1:
The system performs preliminary actions by monitoring network parameters continuously and predicting potential SLA violations before they occur. The machine learning model analyzes historical data and network conditions to forecast future SLA status, enabling proactive intervention rather than reactive response after violations are detected.
Solution Approach 2:
The system implements feedback mechanisms where network performance data is continuously collected, analyzed by machine learning models, and used to adjust routing decisions in real-time. This closed-loop feedback enables dynamic adaptation to changing network conditions, allowing the system to prevent SLA violations before they impact user experience.
2Adaptability or versatility
If network administrators add new applications to the network without extensive testing, then the network can support more applications, but unforeseen issues arise due to assumptions about usage and impact on existing applications
Solution Approach 1:
The system performs preliminary analysis by using machine learning models to predict the impact of new applications on existing network infrastructure before deployment. The model analyzes historical data from similar applications to forecast bandwidth requirements, potential SLA violations, and interactions with existing applications, enabling informed deployment decisions without extensive testing.
Solution Approach 2:
The system creates virtual copies or models of network conditions and application behaviors through machine learning simulations. By replicating network scenarios and analyzing predicted outcomes, administrators can assess potential impacts before actual deployment, reducing the need for extensive physical testing while maintaining network stability.
3Device complexity
If the network routes application flows reactively after SLA violations occur, then the network can maintain simple routing logic, but the application experience deteriorates during the detection and re-routing period
Solution Approach 1:
The system implements preliminary routing actions by predicting potential SLA violations before they occur and proactively adjusting routes in advance. The machine learning model analyzes network conditions and application requirements to forecast future SLA status, enabling the system to prepare alternative routes and switch traffic proactively, thereby maintaining application experience quality without requiring complex reactive logic.
Data Source
AI summary
This disclosure describes techniques and mechanisms for intelligently and proactively routing application traffic within a SD-WAN. The techniques may utilize application models to determine bandwidth utilization and usage patterns of application flow(s), determine and predict the impact of potential movement or addition of the application traffic to a pathway and the SLA requirements of the pathway, and intelligently route and/or re-route the packets through the most optimal path, thereby preventing and/or minimizing policy violations. This disclosure describes techniques and mechanisms for intelligently analyze a target network and determine the impact of adding that application on the network based on usage patterns, flow density, bandwidth requirements and volume of traffic, thereby streamlining the process of designing a new network or adding an application to an existing network.