Network Congestion Management via Dynamic Rate Adjustment

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current network management systems fail to effectively monitor and manage congestion in a granular fashion, such as per link, per connection, or per class of service, and often lack the ability to provide end-to-end delay analysis and appropriate congestion control measures.

Innovation Solution

The system receives congestion indicators and delay information, adjusting data transfer rates based on effective bandwidth determinations, implementing congestion control processes when thresholds are exceeded, and dynamically modifying ingress and egress data rates to manage network performance.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If network monitoring systems are implemented to detect congestion, then network management capability is improved, but the systems fail to provide granular analysis per link, per connection, or per class of service

Engineering Contradiction:
Improvecongestion analysis granularityVSAvoidmonitoring system complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent segments network monitoring into multiple hierarchical levels: link-level monitoring using ECN bits in individual packets, connection-level aggregation at network points, and class-of-service grouping. This segmentation enables granular congestion detection per link, per connection, and per service class without requiring a single complex monolithic system, thereby improving measurement precision while managing device complexity through modular deployment.

Inventive Principle:
Principle #1Segmentation

2Reliability

If data transfer rate is reduced in response to congestion indicators, then network congestion is controlled, but network utilization efficiency decreases

Engineering Contradiction:
Improvecongestion control effectivenessVSAvoidnetwork utilization
Core Design Contradiction:
ReliabilityVSProductivity

Solution Approach 1:

The patent implements dynamic rate adjustment where the data transfer rate is not statically reduced but adaptively modified based on real-time congestion conditions. The system dynamically selects from multiple rate reduction factors (e.g., reducing to effective bandwidth, committed information rate, or other factors) depending on the severity and persistence of congestion, allowing the network to maintain high utilization during light congestion while ensuring reliable congestion control during severe conditions.

Inventive Principle:
Principle #15Dynamics

Solution Approach 2:

The system continuously monitors congestion indicators (such as ECN bits in incoming packets) and uses this feedback to adjust data transfer rates. The feedback loop includes observing congestion over time periods, determining whether congestion is persistent or transient, and accordingly selecting appropriate rate adjustment strategies. This feedback mechanism ensures that rate reduction is applied only when necessary, maintaining network utilization efficiency while achieving reliable congestion control.

Inventive Principle:
Principle #23Feedback

3Reliability

If multiple congestion control processes are implemented for different delay scenarios, then congestion management effectiveness is improved, but system complexity increases

Engineering Contradiction:
Improvecongestion management effectivenessVSAvoidcontrol process complexity
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The patent applies different congestion control processes locally tailored to specific delay scenarios: first congestion control process for when second delay information is received within a threshold time period, and second congestion control process for when it exceeds the threshold. Each process is optimized for its specific condition, and the system automatically selects the appropriate process based on local delay characteristics, improving congestion management effectiveness while keeping overall system complexity manageable through conditional branching rather than universal complexity.

Inventive Principle:
Principle #3Local quality

Data Source

PatentEP2869516B1Systems and methods for managing a network
Publication Date: 2022.04.13 COMCAST CABLE COMM LLC
  • EP2869516B1 patent drawingFigure 1A
  • EP2869516B1 patent drawingFigure 1B~1D
  • EP2869516B1 patent drawingFigure 2

AI summary

Systems and methods for managing congestion in a network are disclosed. One method can comprise receiving a first congestion indicator at a network point and modifying a data transfer rate to an effective bandwidth in response to receiving the first congestion indicator. If a second congestion indicator is received within a predetermined time period, the data transfer rate can be reduced to a factor of a committed information rate. If a second congestion indicator is not received with the time period, the data transfer rate can be increased to a target transfer rate.