Common Problems in Delay-Based Congestion Control Algorithms: A Gallery of Solutions

TL;DR

This paper analyzes the challenges faced by delay-based congestion control algorithms like FAST, focusing on fairness, measurement errors, and coexistence issues, and discusses recent proposals to improve their performance and deployment viability.

Contribution

The paper provides a comprehensive analysis of problems in delay-based congestion control and reviews recent solutions addressing fairness, measurement errors, and coexistence challenges.

Findings

Delay-based algorithms struggle with fairness against loss-based protocols.

Measurement errors in queuing delay impact efficiency and fairness.

Recent proposals offer potential improvements for delay-based congestion control.

Abstract

Although delay-based congestion control protocols such as FAST promise to deliver better performance than traditional TCP Reno, they have not yet been widely incorporated to the Internet. Several factors have contributed to their lack of deployment. Probably, the main contributing factor is that they are not able to compete fairly against loss-based congestion control protocols. In fact, the transmission rate in equilibrium of delay-based approaches is always less than their fair share when they share the network with traditional TCP-Reno derivatives, that employ packet losses as their congestion signal. There are also other performance impairments caused by the sensitivity to errors in the measurement of the congestion signal (queuing delay) that reduce the efficiency and the intra-protocol fairness of the algorithms. In this paper we report, analyze and discuss some recent proposals in the literature to improve the dynamic behavior of delay-based congestion control algorithms, and FAST in particular. Coexistence of sources reacting differently to congestion, identifying congestion appearance in the reverse path and the persistent congestion problem are the issues specifically addressed.