Levelwise Mesh Sparsification for Shortest Path Queries

Shortest path query is a problem to answer the shortest path from the origin to the destination in short time. The network is fixed, and the goal of the study is to develop efficient data structures to speed up the search algorithms which can be constructed in not so long time. In this paper, we propose levelwise mesh sparsification method for constructing a sparse network. The shortest path can be solved in the sparse network, thus the computation time for each query is reduced. We consider regions of several sizes, and construct the sparsified network for each region composed of edges which are parts of shortest paths of vertices far from the region. For each query, the sparse network is constructed by combining the sparsified networks for which the origin and the destination are distant. We show that the sparsified networks are actually very sparse compared to the original network by some computational experiments on real road networks.

• Type: Journal
• Year: 2010
• Where: ISAAC
• Authors: Yuichiro Miyamoto, Takeaki Uno, Mikio Kubo
• Keywords: Algorithms, ISAAC 2010, Shortest Paths, Sparse Network, Sparsified Networks