Scheduled System Maintenance:
On Monday, April 27th, IEEE Xplore will undergo scheduled maintenance from 1:00 PM - 3:00 PM ET (17:00 - 19:00 UTC). No interruption in service is anticipated.
By Topic

BC-GA: A Graph Partitioning Algorithm for Parallel Simulation of Internet Applications

Sign In

Cookies must be enabled to login.After enabling cookies , please use refresh or reload or ctrl+f5 on the browser for the login options.

Formats Non-Member Member
$31 $13
Learn how you can qualify for the best price for this item!
Become an IEEE Member or Subscribe to
IEEE Xplore for exclusive pricing!
close button

puzzle piece

IEEE membership options for an individual and IEEE Xplore subscriptions for an organization offer the most affordable access to essential journal articles, conference papers, standards, eBooks, and eLearning courses.

Learn more about:

IEEE membership

IEEE Xplore subscriptions

2 Author(s)
Siming Lin ; Chinese Acad. of Sci., Beijing ; Xueqi Cheng

Static task mapping of parallel simulation is studied as a graph partitioning problem (GPP). However, the existing algorithms are primarily designed for partitioning finite element meshes or random graphs which are essentially different from the Internet-like topologies used in the field of large scale network simulation. In this paper, we present a new genetic algorithm, BC-GA, for effectively partitioning Internet-like topologies based on boundary crossing, a quite different principle inspired by the analysis of characteristic of the Internet topology and its related solutions. All operations of this algorithm are novel, such as pizza-cutting and autogamy propagation. We test this algorithm on a large extent of graphs, including the snapshots of the real AS-level Internet, the topologies produced by the Internet model generator and many traditional benchmark graphs. The experiment shows that our algorithm can outperform traditional approaches on partitioning Internet-like topologies and it is also better than or comparable to those on traditional GPP. The experiment also shows that a genetic algorithm can converge quickly if the domain knowledge is fitly combined.

Published in:

Parallel, Distributed and Network-Based Processing, 2008. PDP 2008. 16th Euromicro Conference on

Date of Conference:

13-15 Feb. 2008