Distributed transportation simulation is an important technology for evaluating various Intelligent Transportation Systems (ITS) applications, before they are implemented in the real-world traffic system. Offline road network partitioning is the first step towards distributed transportation simulation. However, as the most popular offline road network partitioning solution, METIS cannot naturally formalize data distribution in ITS applications, and thus cannot guarantee to minimize data exchanges between partitions. In this paper, we propose to formalize offline road network partitioning as a hyper graph partitioning, which can naturally deal with data distribution in ITS applications. Then, we propose to solve the hyper graph partitioning using hMETIS, a graph partitioning algorithm borrowed from VLSI applications. Preliminary experiments show that even in the case where there is no ITS application, hyper graph-based offline road network partitioning can reduce data exchanges between partitions by around 10% on average. Currently, we are evaluating hyper graph-based offline road network partitioning on ITS applications.
[1]
Kai Nagel,et al.
Parallel implementation of the TRANSIMS micro-simulation
,
2001,
Parallel Comput..
[2]
V. Voloshin.
Introduction to Graph and Hypergraph Theory
,
2013
.
[3]
Shashi Shekhar,et al.
Multilevel hypergraph partitioning: applications in VLSI domain
,
1999,
IEEE Trans. Very Large Scale Integr. Syst..
[4]
Vipin Kumar,et al.
Multilevel k-way hypergraph partitioning
,
1999,
DAC '99.
[5]
Shashi Shekhar,et al.
Multilevel hypergraph partitioning: application in VLSI domain
,
1997,
DAC.
[6]
Vipin Kumar,et al.
A Fast and High Quality Multilevel Scheme for Partitioning Irregular Graphs
,
1998,
SIAM J. Sci. Comput..
[7]
Moshe Ben-Akiva,et al.
Scalability of dynamic traffic assignment
,
2009
.
[8]
David Charypar,et al.
Efficient algorithms for the microsimulation of travel behavior in very large scenarios
,
2008
.