SDN Flow Admission Control Using Heatmap Saturation Prediction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Flow admission control and network fabric saturation remain challenging in Software Defined Networking (SDN) fabrics, as admitting a flow requiring excessive resources can lead to saturation, causing insufficient resources for all flows, despite centralized control offering better performance over decentralized architectures.
Innovation Solution
A supervisory device predicts the characteristics of new traffic flows using a heatmap-based saturation model and machine learning, adjusting Call Admission Control (CAC) parameters based on telemetry data to proactively manage resource allocation and prevent saturation, employing reinforcement learning to optimize flow admission decisions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If flow admission control is implemented in SDN fabric, then network resource utilization is improved, but system complexity increases
Solution Approach 1:
The patent introduces a heatmap-based saturation model as an intermediary mechanism that visually represents network fabric saturation levels. This model acts as a mediator between flow admission control decisions and actual network resource allocation, simplifying the complex decision-making process by providing intuitive saturation indicators that guide admission control without requiring complex real-time calculations at each decision point
Solution Approach 2:
The system performs preliminary saturation assessment using the heatmap model before admitting new flows to the SDN fabric. By evaluating potential saturation conditions in advance and adjusting CAC parameters proactively, the system prevents resource exhaustion before it occurs, improving resource utilization while maintaining manageable complexity through predictive rather than reactive control
2Measurement precision
If CAC parameters are adjusted dynamically, then flow admission control accuracy is improved, but control complexity increases
Solution Approach 1:
The patent implements a feedback mechanism where the heatmap-based saturation model continuously monitors network fabric utilization and feeds this information back to adjust CAC parameters dynamically. This feedback loop enables accurate flow admission control by adapting to changing network conditions while maintaining reasonable control complexity through automated parameter adjustment based on observed saturation patterns rather than manual tuning
Solution Approach 2:
The system employs reinforcement learning algorithms that enable the flow admission control mechanism to self-adjust CAC parameters based on observed network behavior and saturation patterns. This self-service capability improves control accuracy over time as the system learns from experience, while reducing the need for complex external control intervention and manual parameter tuning
Data Source
AI summary
In one embodiment, a supervisory device for a software defined networking (SDN) fabric predicts characteristics of a new traffic flow to be admitted to the fabric, based on a set of initial packets of the flow. The supervisory device predicts an impact of admitting the flow to the SDN fabric, using a heatmap-based saturation model for the SDN fabric. The supervisory device admits the flow to the SDN fabric, based on the predicted impact. The supervisory device uses reinforcement learning to adjust one or more call admission control (CAC) parameters of the SDN fabric, based on captured telemetry data regarding the admitted flow.