Design and performance evaluation of a state-space based AQM

TL;DR

This paper presents a novel AQM design based on state-space control theory, modeling TCP congestion as a feedback control system, and evaluates its performance through simulations.

Contribution

It introduces a state-space based AQM design using state feedback synthesis for delay systems, linking congestion control with control theory.

Findings

The proposed AQM effectively manages network congestion.

Simulation results demonstrate improved performance under cross traffic.

The control-based approach offers a systematic design method.

Abstract

Recent research has shown the link between congestion control in communication networks and feedback control system. In this paper, the design of an active queue management (AQM) which can be viewed as a controller, is considered. Based on a state space representation of a linearized fluid flow model of TCP, the AQM design is converted to a state feedback synthesis problem for time delay systems. Finally, an example extracted from the literature and simulations via a network simulator NS (under cross traffic conditions) support our study.