Automated Boolean Learning for SDN Deployment Models
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Modeling deployments of policies in software-defined networks (SDNs) is complex and time-consuming, requiring deep understanding of SDN behavior and multiple interactions with engineering teams, and different versions of SDNs exhibit different behaviors, necessitating frequent remodeling.
Innovation Solution
A method and system for generating deployment models for deploying rules in networks by identifying candidate deployment configuration variables and corresponding policy configurations, and determining rule configuration states to formulate a deployment model for implementing policies in response to contract deployments, allowing for automated or semi-automated generation of deployment models.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If manual modeling of policy deployments is performed to ensure accurate representation of SDN behavior, then model accuracy and reliability are improved, but the time and effort required increases significantly
Solution Approach 1:
The system performs self-service by automatically generating deployment models through automated Boolean learning. The model learns and infers deployment behavior patterns from observed SDN controller and switch interactions without requiring manual engineering input, thereby maintaining accuracy while eliminating the time-consuming manual modeling process.
Solution Approach 2:
The patent replaces the manual mechanical process of modeling with an automated computational system. The automated Boolean learning system uses machine learning algorithms to infer deployment models from observed data, substituting human engineering efforts with an automated computational approach that maintains reliability while significantly reducing time investment.
2Adaptability or versatility
If frequent remodeling is performed to adapt to different SDN versions, then adaptability is improved, but the complexity and effort of the process increases
Solution Approach 1:
The system embraces dynamics by continuously learning from observed SDN behavior across different versions. Rather than creating static models for each version, the automated learning system dynamically adapts to new versions by observing their specific behaviors and updating the deployment model accordingly, reducing the complexity of manual remodeling while maintaining high adaptability.
Solution Approach 2:
The patent applies parameter changes by allowing the deployment model to automatically adjust its parameters and rules based on observed behavior from different SDN versions. The automated learning system modifies the Boolean rules and deployment patterns to match the specific characteristics of each SDN version, enabling adaptability without requiring complex manual reconfiguration.
3Measurement precision
If deep understanding of SDN behavior is required to create accurate deployment models, then model precision is improved, but the difficulty of operation increases
Solution Approach 1:
The system implements feedback by continuously observing the actual deployment behavior of SDN controllers and comparing it with the learned model. This feedback loop allows the automated learning system to refine its understanding of SDN behavior and improve modeling precision over time without requiring operators to have deep expert knowledge, thereby maintaining precision while improving ease of operation.
Solution Approach 2:
The automated Boolean learning system acts as an intermediary between the complex SDN behavior and the deployment model. It captures and processes the intricate SDN controller-switch interactions, transforming them into understandable Boolean rules that can be used for deployment modeling, thereby achieving precision without requiring direct human expertise in SDN internals.
Data Source
AI summary
Systems, methods, and computer-readable media for determining a deployment model for deploying rules in a network environment in response to deployment of a contract into the network environment. In some embodiments, a method can include deploying a contract into a network environment. One or more candidate deployment configuration variables can be selected and policy configurations for deploying rules in the network environment as part of implementing policies using the contract can be determined based on the one or more candidate deployment configuration variables. Rule configuration states corresponding to the policy configuration states in the network environment can be identified based on deployment of the one or more contracts in the network environment. Subsequently, a deployment model for implementing the one or more policies in the network environment can be formed based on the policy configurations and the rule configuration states corresponding to the policy configurations.