Virtual WAN Hub Topology Using Latency-Based Placement
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing virtual WAN designs often result in inflated latency and inefficient resource usage due to sub-optimal placement of virtual hubs, leading to higher costs and performance issues for enterprise clients, as they rely on geographic proximity rather than measured latency and network dynamics.
Innovation Solution
A system and method for optimizing virtual WAN topologies by using active latency measurements and a mixed integer linear program (MILP) to select the number and locations of virtual WAN hubs, minimizing weighted latency and number of hubs, and providing a Pareto optimal frontier for trade-offs between latency and cost.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If virtual WAN hubs are placed based on geographic proximity to enterprise clients, then deployment simplicity is improved, but client latency increases due to sub-optimal routing paths
Solution Approach 1:
The patent changes the selection parameter from geographic proximity to measured latency values. The system actively measures latency from multiple entry points to the WAN and uses these measurements to determine optimal hub locations, thereby resolving the contradiction between deployment simplicity and client latency.
Solution Approach 2:
The patent implements a feedback mechanism where latency measurements are continuously obtained and used to adjust hub placement decisions. This feedback loop allows the system to adapt to actual network conditions rather than relying solely on geographic assumptions, reducing client latency while maintaining operational simplicity.
2Loss of time
If more virtual WAN hubs are deployed to reduce latency, then client latency is reduced, but infrastructure cost and complexity increase
Solution Approach 1:
The patent applies partial action by deploying the minimum necessary number of hubs to achieve acceptable latency performance. Rather than excessively deploying hubs throughout the infrastructure, the system uses latency measurements to identify and deploy hubs only where they provide the most benefit, thus reducing both latency and infrastructure complexity.
Solution Approach 2:
The patent implements dynamic hub placement based on measured latency conditions rather than static geographic assumptions. The system can adapt hub locations and numbers based on actual network performance data, allowing optimal balance between latency reduction and infrastructure complexity without over-provisioning.
3Adaptability or versatility
If virtual WAN hubs are placed at geographically distributed locations, then service coverage is improved, but resource usage efficiency decreases due to sub-optimal routing
Solution Approach 1:
The patent changes the location selection parameter from geographic distribution to latency-based optimization. By measuring actual latency from entry points to the WAN, the system can place hubs at locations that provide both wide service coverage and efficient resource usage, resolving the contradiction between coverage and efficiency.
Solution Approach 2:
The patent performs preliminary latency measurements and analysis before deploying virtual WAN hubs. This preliminary action allows the system to pre-determine optimal hub locations that will achieve both wide service coverage and efficient resource usage, avoiding sub-optimal routing and wasted resources.
4Quantity of substance
If enterprise clients rely on public Internet routing, then infrastructure investment is reduced, but routing control and performance are lost
Solution Approach 1:
The patent introduces virtual WAN hubs as intermediary points that enterprises can deploy through cloud providers. These hubs act as mediators between the enterprise and the public Internet infrastructure, providing routing control and performance optimization without requiring enterprises to build their own dedicated WAN infrastructure.
Solution Approach 2:
The patent enables enterprises to copy the routing control capabilities of dedicated WANs through virtualized hub deployments. By deploying virtual hubs through cloud providers, enterprises can replicate the performance and control characteristics of private WAN infrastructure without the significant capital investment required for building dedicated physical infrastructure.
Data Source
AI summary
The present application relates to a system, apparatus, and method of selecting a topology for a virtual wide area network (vWAN). A wide area network (WAN) includes a plurality of geographically distributed entry points and a plurality of edge datacenters. Each edge datacenter is associated with an entry point and includes computing resources for hosting a vWAN hub. A vWAN manager is configured to: obtain latency measurements from different metropolitan regions to the entry points; receive a set of the metropolitan regions and a number of expected connections for each metropolitan region; and select a plurality of vWAN hub locations at selected edge datacenters, each selected edge datacenter being associated with a non-empty subset of the metropolitan regions. A weighted latency for the plurality of vWAN hub locations based on the latency measurements and the expected connections is minimized for a number of the selected edge datacenters.