Adaptive Network Rate Control via Congestion Signal Correlation

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing rate control techniques in data networks often fail to achieve high network bandwidth utilization, fair bandwidth allocation, and dynamic adaptation to congestion, especially in the presence of aggressive flows, and are not resilient to noise and random packet loss, leading to slow convergence and oscillations.

Innovation Solution

The implementation of adaptive learning techniques that utilize congestion signals such as packet delay, loss, and ECN marking to dynamically adjust transmission rates, with a learning module correlating these signals with actual congestion levels and adjusting the desired operating congestion level to maintain network efficiency and fairness.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If traditional rate control algorithms are used, then network bandwidth utilization is improved, but fairness among multiple flows deteriorates

Engineering Contradiction:

Improvenetwork bandwidth utilizationVSAvoidfairness among flows

Core Design Contradiction:

ProductivityVSReliability

Solution Approach 1:

The patent implements a feedback mechanism where the rate control algorithm continuously monitors network congestion signals (packet loss, delay, ECN markings) and adjusts transmission rates based on this feedback. This allows the system to dynamically respond to changing network conditions while maintaining fairness through observed congestion levels rather than aggressive contention

Inventive Principle:

Principle #23Feedback

Solution Approach 2:

The patent changes the control parameters from fixed rate control to adaptive rate control based on observed congestion signals. The transmission rate is adjusted as a function of observed packet loss, delay, and ECN markings, allowing the system to adapt to different network conditions and maintain both high utilization and fairness

Inventive Principle:

Principle #35Parameter changes

2Productivity

If aggressive rate control is used to achieve high bandwidth utilization, then network efficiency is improved, but convergence speed to steady state deteriorates

Engineering Contradiction:

Improvenetwork efficiencyVSAvoidconvergence time to steady state

Core Design Contradiction:

ProductivityVSLoss of time

Solution Approach 1:

The patent implements dynamic rate control where the transmission rate continuously adapts to observed congestion conditions. The rate control is dynamic rather than static, allowing the system to converge to steady state more quickly by responding to actual network conditions rather than using fixed aggressive rates

Inventive Principle:

Principle #15Dynamics

Solution Approach 2:

The feedback mechanism allows the system to quickly respond to congestion signals and adjust rates accordingly, achieving faster convergence to steady state compared to traditional algorithms that use fixed rate control patterns

Inventive Principle:

Principle #23Feedback

3Stability of the object's composition

If fixed congestion level control is used, then network stability is improved, but adaptability to changing network conditions deteriorates

Engineering Contradiction:

Improvenetwork stabilityVSAvoidadaptability to congestion changes

Core Design Contradiction:

Stability of the object's compositionVSAdaptability or versatility

Solution Approach 1:

The patent implements dynamic congestion level control where the desired congestion level and control parameters adapt to changing network conditions. The system maintains stability through controlled adaptation rather than fixed parameters, allowing it to respond to varying congestion levels and network states

Inventive Principle:

Principle #15Dynamics

4Device complexity

If simple rate control algorithms are used, then device complexity is reduced, but resilience to noise and random packet loss deteriorates

Engineering Contradiction:

Improvealgorithm complexityVSAvoidresilience to noise and packet loss

Core Design Contradiction:

Device complexityVSReliability

Solution Approach 1:

The patent uses feedback from multiple congestion signals (packet loss, delay, ECN markings) to improve resilience to noise and random packet loss. By observing multiple indicators rather than relying on a single signal, the system can distinguish between congestion-induced loss and random loss, maintaining reliability without excessive complexity

Inventive Principle:

Principle #23Feedback

Data Source

PatentUS8570864B2Kernel awareness of physical environment

Publication Date: 2013.10.29 MICROSOFT TECHNOLOGY LICENSING LLC

AI summary

Described are techniques to use adaptive learning to control bandwidth or rate of transmission of a computer on a network. Congestion observations such as packet delay and packet loss are used to compute a congestion signal. The congestion signal is correlated with information about actual congestion on the network, and the transmission rate is adjusted according to the degree of correlation. Transmission rate may not adjust when packet delay or packet loss is not strongly correlated with actual congestion. The congestion signal is adaptively learned. For instance, the relative effects of loss and delay on the congestion signal may change over time. Moreover, an operating congestion level may be minimized by adaptive adjustment.