SD-WAN Tunnel Peering for Predictive SLA-Aware Traffic Rerouting
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing SD-WAN tunnel configurations are static and often unused, leading to inefficiencies and SLA failures due to the dynamic nature of modern network traffic, which relies on outdated proxies for user experience and lacks predictive routing to optimize application performance.
Innovation Solution
A predictive application aware routing engine uses machine learning and telemetry to identify available SD-WAN tunnels across enterprises, allowing dynamic peering and rerouting to optimize application experience by predicting and preventing SLA violations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If static SD-WAN tunnel configurations are used, then network security and SLA guarantees are maintained, but tunnel utilization is low and traffic efficiency deteriorates
Solution Approach 1:
The patent implements dynamic tunnel configuration where the SD-WAN controller continuously monitors traffic patterns, tunnel performance, and resource availability, then automatically adjusts tunnel routing decisions in real-time based on current conditions rather than relying on static pre-configured paths
Solution Approach 2:
The system employs feedback mechanisms where performance metrics from active tunnels are continuously collected and fed back to the controller, which uses this information to make intelligent routing decisions about traffic distribution across available tunnels, optimizing utilization while maintaining SLA compliance
2Quantity of substance
If more SD-WAN tunnels are configured to handle dynamic traffic, then traffic capacity increases, but most tunnels remain unused and resource waste occurs
Solution Approach 1:
The patent enables tunnels to serve multiple purposes and multiple enterprises simultaneously. A single tunnel infrastructure can dynamically carry traffic for different SD-WAN networks based on demand, allowing underutilized tunnels to be shared and repurposed for other traffic flows, thereby increasing overall capacity utilization without requiring additional dedicated tunnels for each enterprise
Solution Approach 2:
The system dynamically changes tunnel parameters such as routing assignments, traffic allocation weights, and activation states based on real-time monitoring of traffic patterns and tunnel utilization metrics, allowing the network to adapt capacity allocation to actual demand rather than maintaining fixed allocations
3Measurement precision
If traditional KPI thresholds are used to monitor network performance, then SLA compliance can be tracked, but predictive routing capability is lost and user experience degradation occurs
Solution Approach 1:
The patent implements predictive analytics that analyze historical traffic patterns, tunnel performance trends, and external factors to forecast future SLA violations or performance degradation before they occur. This allows the system to proactively reroute traffic or adjust parameters to prevent user experience degradation rather than merely detecting it after thresholds are breached
Data Source
AI summary
In one embodiment, a device identifies available resources of a tunnel in a first software defined network. The device provides, based on the available resources, an indication that the tunnel is available to convey traffic sent by a second software defined network. The device receives, based on the indication, a request to convey traffic sent by the second software defined network via the tunnel in the first software defined network. The device configures a peering node in the first software defined network to connect the second software defined network to the tunnel to allow the traffic sent by the second software defined network to be conveyed via the tunnel.