Home

About

Advanced Search

Browse by Discipline

Scientific Societies

E-print Alerts

Add E-prints

E-print Network
FAQHELPSITE MAPCONTACT US


  Advanced Search  

 
KYBERNETIKA | VOLUME 3 7 (2001), NUMBER 3, PAGES 355 { 365 STATISTICAL{LEARNING CONTROL
 

Summary: KYBERNETIKA | VOLUME 3 7 (2001), NUMBER 3, PAGES 355 { 365
STATISTICAL{LEARNING CONTROL
OF MULTIPLE{DELAY SYSTEMS
WITH APPLICATIONS TO ATM NETWORKS
C.T. Abdallah1, M. Ariola2 and V. Koltchinskii3
Congestion control in the ABR class of ATM network presents interesting challenges
due to the presence of multiple uncertain delays. Recently, probabilistic methods and
statistical learning theory have been shown to provide approximate solutions to challenging
control problems. In this paper, using some recent results by the authors, an e cient
statistical algorithm is used to design a robust, xed-structure, controller for a high-speed
communication network with multiple uncertain propagation delays.
1. INTRODUCTION
This paper illustrates the application of statistical-learning control results for an
Available Bit Rate (ABR) congestion control algorithm in an Asynchronous Trans-
fer Mode (ATM) communications network. The ABR service categoryis a best-e ort
class used in ATM networks to handle highly bursty and varying data applications.
ATM was selected by the International Telecommunication Union (ITU) for Broad-
band Integrated Service Digital Network (B-ISDN), and is detailed in 2]. ATM
requires the transmission of xed size cells (each containing 53 bytes) and is a
connection-based network combining the advantages of packet and circuit switch-

  

Source: Abdallah, Chaouki T- Electrical and Computer Engineering Department, University of New Mexico

 

Collections: Engineering