Controller Tuning for Active Queue Management Using a Parameter Space Method

TL;DR

This paper presents a method for tuning simple PID controllers for Active Queue Management using a parameter space approach that identifies stability regions based on delay-based linear models.

Contribution

It introduces a novel parameter space method for tuning PID controllers specifically for AQM, considering delay effects in the linear models.

Findings

Effective stability regions identified for PID tuning in AQM

Enhanced controller design for network traffic management

Applicable to delay-based linear models in practice

Abstract

In recent years, different mathematical models have been proposed for widely used internet control mechanisms. Simple low order controllers (such as PID, and Smith predictor based linear controllers that are easy to implement) are desired for network traffic management. In order to design such simple controllers for Active Queue Management (AQM), delay based linear models have been considered. In this paper we discuss tuning of the PID controllers by using a parameter space method, which computes stability regions of a class of quasi-polynomials in terms of free controller parameters.