ML-Based SD-WAN Configuration Optimization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Software-defined wide area networks (SD-WANs) are complex and prone to misconfigurations due to varying transport dynamics, application requirements, and geographical differences, leading to suboptimal configuration settings that can result in service level agreement (SLA) failures and poor Quality of Experience (QoE).
Innovation Solution
A device uses machine learning to associate application performance with network configuration changes across SD-WANs, training a model to predict the effect of configuration changes and generate recommendations for optimizing network settings, thereby improving application performance and reducing SLA violations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If manual configuration methods are used for SD-WAN networks, then device complexity is reduced, but configuration accuracy deteriorates leading to misconfigurations and SLA failures
Solution Approach 1:
The system enables self-service configuration by using machine learning models to automatically generate optimized network configurations based on observed traffic patterns and performance metrics, eliminating the need for manual expert configuration while maintaining high accuracy across diverse network conditions
Solution Approach 2:
The system implements feedback mechanisms where configuration changes are continuously monitored for their impact on SLA compliance and QoE metrics, using this feedback to refine the machine learning models and improve future configuration recommendations, thereby achieving high precision through iterative learning
2Adaptability or versatility
If static configuration settings are used, then ease of operation is improved, but adaptability deteriorates causing suboptimal performance in dynamic network conditions
Solution Approach 1:
The system transitions from static to dynamic configurations by continuously adapting network parameters based on real-time traffic patterns, SLA compliance status, and QoE metrics, allowing the configuration to automatically adjust to changing network conditions while maintaining operational simplicity through automated decision-making
Solution Approach 2:
The machine learning models perform preliminary analysis of network conditions and predict optimal configuration changes before SLA violations occur, proactively adjusting settings to prevent performance degradation and maintain service quality under varying network dynamics
3Measurement precision
If comprehensive monitoring of all configuration parameters is implemented, then measurement precision is improved, but loss of information increases due to the complexity and volume of data
Solution Approach 1:
The system extracts only the critical information needed for configuration optimization by focusing measurement on key SLA parameters (latency, jitter, packet loss) and QoE metrics, filtering out redundant data and concentrating computational resources on analyzing the most impactful parameters for predicting and optimizing network performance
Data Source
AI summary
In one embodiment, a device associates application performance of an online application with network configuration changes implemented across one or more software-defined networks. The device trains a machine learning model to predict an effect of a configuration change on the application performance for any given portion of the one or more software-defined networks. The device generates a recommended configuration change for a particular portion of the one or more software-defined networks, using the machine learning model. The device causes the recommended configuration change to be implemented in the particular portion of the one or more software-defined networks.