Analysis of the ARB (adaptive rate based) congestion avoidance algorithm

Author

Artiges, Damien

Author_Institution

Inst. Nat. de Recherche en Inf. et Autom., Sophia-Antipolis, France

Volume

3

fYear

1996

fDate

11-13 Dec 1996

Firstpage

2899

Abstract

The ARB (adaptive rate based) algorithm is a congestion control mechanism developed by IBM, Research Triangle Park, North Carolina. ARB is a distributed algorithm based on the periodic exchange of information between the two end points of a connection. The sender uses the feedback, in particular a delay estimation, to detect approaching congestion and regulate the input traffic accordingly. We define a discrete time Markov chain modeling a simple network controlled by the ARB algorithm. We show that the sending rate and the queue length are bounded, and we establish the geometric convergence of the system toward steady-state behavior

Keywords

Markov processes; adaptive control; computer networks; delays; discrete time systems; distributed algorithms; distributed control; feedback; performance evaluation; queueing theory; telecommunication congestion control; ARB algorithm; adaptive rate based congestion avoidance algorithm; bounded queue length; bounded sending rate; computer networks; delay estimation; discrete time Markov chain; distributed algorithm; feedback; geometric convergence; periodic information exchange; steady-state behavior; Adaptive control; Algorithm design and analysis; Communication system traffic control; Delay; Distributed algorithms; Feedback; Programmable control; Steady-state; Stochastic processes; Traffic control;

fLanguage

English

Publisher

ieee

Conference_Titel

Decision and Control, 1996., Proceedings of the 35th IEEE Conference on

Conference_Location

Kobe

ISSN

0191-2216

Print_ISBN

0-7803-3590-2

Type

conf

DOI

10.1109/CDC.1996.573561

Filename

573561