Rate-Based PI Controller for TCP Active Queue Management
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traditional queue management mechanisms, such as drop tail and queue-based AQM schemes, are inadequate in managing congestion effectively, especially in networks with small buffers, leading to poor performance and global synchronization of traffic sources.
Innovation Solution
A rate-based proportional-integral control scheme is introduced, which calculates a mark/drop probability based on the error between the assigned capacity and data arrival rate, using a stable gain combination of proportional and integral controller gains, and executes a packet mark/drop routine to control data flow in a data network.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If queue-based AQM schemes are used, then congestion control performance is improved, but sufficient buffering is required which increases device complexity
Solution Approach 1:
The patent changes the control parameter from queue size to data arrival rate. By measuring the data arrival rate and comparing it to the assigned capacity, the system can perform effective congestion control without requiring large buffers, thus resolving the contradiction between performance and buffering requirements
Solution Approach 2:
The patent replaces the mechanical buffering approach with a rate-based control mechanism. Instead of relying on physical queue depth information, the system uses measured arrival rates and control theory to achieve congestion control, eliminating the need for large buffers
2Ease of manufacture
If drop tail mechanism is used, then implementation simplicity is improved, but congestion detection delay increases leading to poor performance
Solution Approach 1:
The patent implements preliminary congestion control by measuring data arrival rates and proactively adjusting mark/drop probabilities before queues become full. This early intervention prevents congestion rather than reacting to it, eliminating the delay inherent in drop tail mechanisms while maintaining implementation simplicity
Solution Approach 2:
The patent introduces a feedback control mechanism where the measured data arrival rate is continuously compared to the assigned capacity, and the mark/drop probability is adjusted based on the error signal. This closed-loop feedback enables timely congestion detection and response without the delays of traditional drop tail approaches
3Measurement precision
If queue size information is used for control, then congestion control accuracy is improved, but small buffers complicate the control problem
Solution Approach 1:
The patent changes the measurement parameter from queue size to data arrival rate. By measuring the rate at which data arrives and comparing it to the assigned capacity, the system achieves accurate congestion control without the complexities associated with small buffer management and queue size-based control
Data Source
AI summary
Disclosed is an Active Queue Management method and apparatus which uses traffic rate information for congestion control. Using a nonlinear fluid-flow model of Traffic Control Protocol, a proportional-integral controller in a closed loop configuration with gain settings characterized for stable operation allows a matching of the aggregate rate of the active TCP connections to the available capacity. Further disclosed is a method for calculation of the regime of gain settings for which stable operation of a given network obtains. This approach allows for capacity matching while maintaining minimal queue size and high link utilization.