The effect of time delays on the stability of load balancing algorithms for parallel computations

John Chiasson; Zhong Tang; Jean Ghanem; Chaouki T. Abdallah; J. Douglas Birdwell; Majeed M. Hayat; Henry Jérez

doi:10.1109/TCST.2005.854339

The effect of time delays on the stability of load balancing algorithms for parallel computations

John Chiasson, Zhong Tang, Jean Ghanem, Chaouki T. Abdallah, J. Douglas Birdwell, Majeed M. Hayat, Henry Jérez

Electrical Engineering Department

Research output: Contribution to journal › Article › peer-review

34 Scopus citations

Abstract

A deterministic dynamic nonlinear time-delay system is developed to model load balancing in a cluster of computer nodes used for parallel computations. The model is shown to be self- consistent in that the queue lengths cannot go negative and the total number of tasks in all the queues and the network are conserved (i.e., load balancing can neither create nor lose tasks). Further, it is shown that using the proposed load balancing algorithms, the system is stable in the sense of Lyapunov. Experimental results are presented and compared with the predicted results from the analytical model. In particular, simulations of the models are compared with an experimental implementation of the load balancing algorithm on a distributed computing network.

Original language	English
Pages (from-to)	932-942
Number of pages	11
Journal	IEEE Transactions on Control Systems Technology
Volume	13
Issue number	6
DOIs	https://doi.org/10.1109/TCST.2005.854339
State	Published - Nov 2005

Keywords

Computer networks
Load balancing
Time delay systems

Access to Document

10.1109/TCST.2005.854339

Cite this

@article{de459eb828b44aa58f58675bf39c0048,

title = "The effect of time delays on the stability of load balancing algorithms for parallel computations",

abstract = "A deterministic dynamic nonlinear time-delay system is developed to model load balancing in a cluster of computer nodes used for parallel computations. The model is shown to be self- consistent in that the queue lengths cannot go negative and the total number of tasks in all the queues and the network are conserved (i.e., load balancing can neither create nor lose tasks). Further, it is shown that using the proposed load balancing algorithms, the system is stable in the sense of Lyapunov. Experimental results are presented and compared with the predicted results from the analytical model. In particular, simulations of the models are compared with an experimental implementation of the load balancing algorithm on a distributed computing network.",

keywords = "Computer networks, Load balancing, Time delay systems",

author = "John Chiasson and Zhong Tang and Jean Ghanem and Abdallah, {Chaouki T.} and Birdwell, {J. Douglas} and Hayat, {Majeed M.} and Henry J{\'e}rez",

year = "2005",

month = nov,

doi = "10.1109/TCST.2005.854339",

language = "English",

volume = "13",

pages = "932--942",

number = "6",

}

TY - JOUR

T1 - The effect of time delays on the stability of load balancing algorithms for parallel computations

AU - Chiasson, John

AU - Tang, Zhong

AU - Ghanem, Jean

AU - Abdallah, Chaouki T.

AU - Birdwell, J. Douglas

AU - Hayat, Majeed M.

AU - Jérez, Henry

PY - 2005/11

Y1 - 2005/11

N2 - A deterministic dynamic nonlinear time-delay system is developed to model load balancing in a cluster of computer nodes used for parallel computations. The model is shown to be self- consistent in that the queue lengths cannot go negative and the total number of tasks in all the queues and the network are conserved (i.e., load balancing can neither create nor lose tasks). Further, it is shown that using the proposed load balancing algorithms, the system is stable in the sense of Lyapunov. Experimental results are presented and compared with the predicted results from the analytical model. In particular, simulations of the models are compared with an experimental implementation of the load balancing algorithm on a distributed computing network.

AB - A deterministic dynamic nonlinear time-delay system is developed to model load balancing in a cluster of computer nodes used for parallel computations. The model is shown to be self- consistent in that the queue lengths cannot go negative and the total number of tasks in all the queues and the network are conserved (i.e., load balancing can neither create nor lose tasks). Further, it is shown that using the proposed load balancing algorithms, the system is stable in the sense of Lyapunov. Experimental results are presented and compared with the predicted results from the analytical model. In particular, simulations of the models are compared with an experimental implementation of the load balancing algorithm on a distributed computing network.

KW - Computer networks

KW - Load balancing

KW - Time delay systems

UR - http://www.scopus.com/inward/record.url?scp=27844581294&partnerID=8YFLogxK

U2 - 10.1109/TCST.2005.854339

DO - 10.1109/TCST.2005.854339

M3 - Article

AN - SCOPUS:27844581294

SN - 1063-6536

VL - 13

SP - 932

EP - 942

JO - IEEE Transactions on Control Systems Technology

JF - IEEE Transactions on Control Systems Technology

IS - 6

ER -

The effect of time delays on the stability of load balancing algorithms for parallel computations

Abstract

Keywords

Access to Document

Other files and links

Fingerprint

Cite this