Packet Network Congestion Control for Real-Time Optimization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing network performance optimization methods are inefficient due to the large number of possible congestion control (CC) scheme and parameter combinations, lacking visibility into how network performance changes with reconfiguration, leading to suboptimal network performance in data centers using RoCE protocol.
Innovation Solution
A system and method for real-time selection and tuning of CC schemes and parameters in network nodes, using graphical display of performance indicators to assist administrators in iteratively optimizing network performance by selecting and adjusting CC schemes and parameters based on visual feedback.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If multiple congestion control schemes and parameter combinations are configured in network nodes, then network performance optimization potential increases, but the complexity of selection and configuration increases
Solution Approach 1:
The system continuously monitors network performance metrics (throughput, latency, packet loss) and uses this feedback to automatically evaluate and select optimal congestion control schemes and parameter combinations, eliminating the need for manual configuration complexity while maintaining performance optimization
Solution Approach 2:
Network nodes autonomously select and adjust their own congestion control parameters based on real-time network conditions and performance feedback, without requiring external configuration or intervention, thereby reducing configuration complexity while achieving performance optimization
2Productivity
If congestion control schemes are reconfigured to improve network performance, then performance metrics improve, but the time required for optimization increases
Solution Approach 1:
The system pre-evaluates multiple congestion control schemes and parameter combinations offline to create a library of optimized configurations, allowing rapid deployment of pre-tested optimal settings without requiring time-consuming real-time experimentation
Solution Approach 2:
Real-time performance monitoring provides immediate feedback on network conditions, enabling the system to quickly identify when reconfiguration is needed and select the optimal scheme from pre-evaluated options, minimizing the time loss during optimization transitions
3Adaptability or versatility
If manual selection and tuning of congestion control parameters is performed, then flexibility in optimization is improved, but the ease of operation decreases
Solution Approach 1:
The system automatically performs congestion control scheme selection and parameter tuning based on real-time network conditions and performance metrics, eliminating the need for manual intervention while maintaining the flexibility to adapt to different network scenarios through algorithmic decision-making
Data Source
AI summary
A communication system (20) includes a plurality of Network Interface Controllers (NICs) (32) and one or more processors (28, 62, 64). The plurality of NICs are to connect multiple hosts to a communication network, the NICs supporting a configurable Congestion Control (CC) scheme (80) selected from among multiple CC schemes. The one or more processors are coupled to the communication network, and are to receive performance indicators indicative of congestion states occurring in the communication network due to communication of the hosts with one another over the communication network, the performance indicators being associated with respective times of occurrence, select respective CC schemes for one or more of the NICs based on the performance indicators and corresponding times of occurrence, and provision the selected CC schemes in the one or more of the NICs.