Tradeoff decisions in the design of a backbone computer network using visualization

Abstract

Visualization provides a useful tool for analyzing large, complex data sets. In the design of backbone computer networks, rough-cut design decisions can gain from a visual analysis of the generated solution with respect to design parameters such as average message delay, delay cost, average message length and total network operating costs. In this paper, we show how two-dimensional and three-dimensional surface and glyph representations can be used for understanding the cost-delay tradeoffs involved in the network design problem, and an idea of the 'efficient frontier' where the user may choose to operate. It provides an opportunity to revisit relationships that exist between the different network design parameters as well as discover new ones. © 2002 Published by Elsevier Science B.V.