SDWAN Load Balancing via QoS-Aware Dynamic Weight Adjustment
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Legacy WAN architectures face challenges such as insufficient bandwidth, high bandwidth costs, application downtime, poor SaaS performance, complex operations, and difficulty in securing networks, due to insufficient dynamic load-balancing across multiple interfaces informed by Quality of Service (QoS) metrics.
Innovation Solution
The dynamic adjustment of load-balancing weights across multiple network transport interfaces based on adaptive Quality of Service (QoS) metrics, using a combination of Adaptive QoS Path Monitor and Software Defined WAN (SDWAN) Weighted Forwarding module, to prevent traffic overruns and maximize available bandwidth usage by rerouting traffic based on real-time bandwidth capacity and utilization.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If static load-balancing weights are used across multiple network interfaces, then device complexity is reduced and operation is simplified, but bandwidth utilization is suboptimal and QoS requirements cannot be dynamically met
Solution Approach 1:
The patent implements dynamic load-balancing weights that automatically adjust based on real-time QoS metrics and bandwidth availability. The system transitions from static to dynamic weight assignment, allowing the load-balancing mechanism to adapt to changing network conditions, thereby maximizing bandwidth utilization while maintaining manageable complexity through automated control
Solution Approach 2:
The system continuously monitors QoS metrics and bandwidth utilization across multiple network interfaces, using this feedback information to dynamically adjust load-balancing weights. This closed-loop control enables the system to optimize bandwidth utilization in response to actual network conditions while maintaining operational simplicity through automated decision-making
2Reliability
If traffic is routed through a single primary interface, then QoS parameters are easier to guarantee, but bandwidth capacity is insufficient and link underutilization occurs
Solution Approach 1:
The patent segments traffic flow across multiple network interfaces rather than concentrating all traffic on a single primary interface. By dividing traffic into different paths with dynamically adjusted weights, the system increases total bandwidth capacity while maintaining QoS guarantees through controlled distribution across segmented paths
Solution Approach 2:
The system creates a composite networking approach by combining multiple interface paths with different characteristics into a unified load-balancing system. This composite structure allows the network to leverage the combined bandwidth capacity of multiple interfaces while maintaining QoS requirements through intelligent traffic distribution
3Productivity
If load-balancing weights are dynamically adjusted based on QoS metrics, then bandwidth utilization is maximized and QoS requirements are met, but system complexity and measurement requirements increase
Solution Approach 1:
The load-balancing system performs self-adjustment by automatically monitoring QoS metrics and modifying its own weight assignments based on observed network conditions. This self-service capability eliminates the need for external manual configuration and reduces operational complexity despite the sophisticated dynamic adjustment mechanisms
Data Source
AI summary
Aspects of the present disclosure are directed to dynamic adjustment of load-balancing weights across multiple network transport interfaces in a network, informed in part by Quality of Service (QoS) metrics. In one aspect, a method includes determining one or more metrics based on one or more Software-defined Wide Area Network (SDWAN) session level throughput and SDWAN session loss through one or more tunnels; generating a Quality of Service (QoS) SDWAN session level shape rate per tunnel based on the one or more metrics; and dynamically adjusting an SDWAN forwarding load-balance weight for each of the one or more tunnels based on the QoS SDWAN session level shape rate.