An efficient parallel optimization algorithm for the Token Bucket control mechanism

Authors:
N. U. Ahmed;X. Lu;L. O. Barbosa
Affiliations:
School in Information Technology and Engineering, University of Ottawa, Ottawa, Ontario, Canada K1N6N5;School in Information Technology and Engineering, University of Ottawa, Ottawa, Ontario, Canada K1N6N5;Instituto de Investigación en Informática de Albacete Universidad de Castilla La Mancha, Campus Universitario s/n, 02071 Albacete, Spain
Venue:
Computer Communications
Year:
2006

Citing 4
Cited 0

An optimal rate control algorithm for guaranteed services in broadband networks

Computer Networks: The International Journal of Computer and Telecommunications Networking
Dynamic Programming

Dynamic Programming
Traffic specifications for the transmission of stored MPEG video onthe Internet

IEEE Transactions on Multimedia
Token bucket characterization of long-range dependent traffic

Computer Communications

Quantified Score

Hi-index	0.25

Visualization

Abstract

The Token Bucket algorithm, one of the most widely used control mechanism in computer communication network, has been extensively used to ensure the QoS needs for various applications. Recently, [N.U. Ahmed, Bo Li, Luis Orozco-Barbosa, Modelling and optimization of computer network traffic controllers, Mathematical Problems in Engineering, 2005, 6(2005), 617-640] an optimization technique, based on dynamic programming and genetic algorithm, has been developed which improves network utilization or throughput by reducing data losses and service time, etc. This, however, requires long execution time and excessive memory space thereby imposing limitation on its applicability to high dimensional problems. In this study we have conserved both space and time complexity. This is achieved by introducing multiple processors and the Reduced Memory Algorithm thereby opening up the prospects of solving large-scale problems. Our parallel processing algorithm is tested with MPEG-4 traces. Our numerical results show that the algorithm can effectively solve the multiple Token Bucket problems. The results also provide us with the guidelines to configure the parallel processing platform.