On the AIMD Algorithm Under Saturation Constraints

One of the most successful distributed resource allocation algorithms to be deployed in industry is the Additive-Increase Multiplicative-Decrease (AIMD) algorithm of Jain and Chiu. This algorithm already underpins the transport layer of the internet, and is now starting to find application in new ar...

Full description

Saved in:
Bibliographic Details
Published in:IEEE transactions on automatic control Vol. 62; no. 12; pp. 6392 - 6398
Main Authors: Studli, S., Corless, M., Middleton, R. H., Shorten, R.
Format: Journal Article
Language:English
Published: IEEE 01.12.2017
Subjects:
AIMD
Artificial intelligence
congestion control
Context awareness
Convergence
distributed control
Heuristic algorithms
positive systems
resource allocation algorithm
Resource management
TCP
Transportation
ISSN:0018-9286, 1558-2523
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:One of the most successful distributed resource allocation algorithms to be deployed in industry is the Additive-Increase Multiplicative-Decrease (AIMD) algorithm of Jain and Chiu. This algorithm already underpins the transport layer of the internet, and is now starting to find application in new areas in the context of Smart Grid and Smart Transportation applications. A distinguishing feature of these latter application areas, when compared with the internet, is the need to modify AIMD to account for lower and upper bounds in the resource allocated to individual agents. Our objective in this note is to demonstrate that the dynamic system arising from AIMD in this extended form has system theoretic properties similar to the classical algorithm.
ISSN:0018-9286
1558-2523
DOI:10.1109/TAC.2017.2651160