Hybrid Predictive Reactive Routing for SD-WAN Failure Handling
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traditional reactive failure detection in SD-WANs is ineffective as it requires actual failure occurrence before rerouting, leading to temporary traffic disruption, whereas predictive failure detection using machine learning offers proactive rerouting but may not cover all failure types, necessitating a hybrid approach.
Innovation Solution
Coupling reactive routing with predictive routing by using machine learning to predict network element failures, updating network topology, recomputing reactive routing tables, and notifying other devices through reactive routing protocol messages, allowing proactive rerouting based on predicted failures.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If reactive failure detection is used, then the network can handle unpredictable failures, but traffic disruption occurs until rerouting is initiated
Solution Approach 1:
The system performs preliminary actions by using machine learning models to predict potential network failures before they actually occur. The predictive routing component analyzes historical data, network patterns, and real-time metrics to forecast failures and proactively reroute traffic, thereby avoiding traffic disruption rather than reacting after disruption has occurred.
Solution Approach 2:
The patent merges reactive routing protocols with predictive routing components into a hybrid system. The reactive component handles unpredictable failures through traditional detection and rerouting mechanisms, while the predictive component uses machine learning to anticipate failures. This combination allows the network to benefit from both approaches: proactive rerouting for predictable failures and reliable fallback for unpredictable ones.
2Loss of time
If predictive routing is used, then traffic disruption is minimized, but not all failure types can be detected
Solution Approach 1:
The hybrid routing system provides universal coverage by combining multiple failure detection approaches. The predictive routing component handles predictable failures through machine learning predictions, while the reactive routing component handles unpredictable failures through traditional detection mechanisms. This multi-functional approach ensures comprehensive failure detection coverage across different failure types.
Solution Approach 2:
The reactive routing protocol serves as an intermediary mechanism that complements the predictive routing system. When the predictive component cannot detect a failure type, the reactive component acts as a mediator to provide traditional failure detection and rerouting capabilities, ensuring continuous network reliability.
3Productivity
If machine learning-based predictive routing is implemented, then proactive rerouting is enabled, but system complexity increases
Solution Approach 1:
The routing system is segmented into distinct functional components: a predictive routing component that uses machine learning for proactive failure detection and rerouting decisions, and a reactive routing component that handles traditional failure detection. This segmentation allows each component to specialize in specific tasks, improving overall rerouting efficiency while managing system complexity through modular architecture.
Data Source
AI summary
In one embodiment, a device in a network predicts failure of a network element in the network using a machine learning-based failure prediction model. The device updates, based on the predicted failure of the network element, a topology of the network to remove the network element from the topology of the network. The device recomputes a reactive routing table of the device using the updated topology of the network. The device notifies one or more other devices of the network of the predicted failure using a reactive routing protocol message.