Indexing of Spatiotemporal Trajectories for Efficient Distance Threshold Similarity Searches on the GPU
Information and Computer Sciences Department, University of Hawai’i at Manoa, Honolulu, HI, U.S.A.
29th IEEE International Parallel & Distributed Processing Symposium (IPDPS), 2015
@article{gowanlock2015indexing,
title={Indexing of Spatiotemporal Trajectories for Efficient Distance Threshold Similarity Searches on the GPU},
author={Gowanlock, Michael and Casanova, Henri},
year={2015}
}
Applications in many domains search moving object trajectory databases. The distance threshold search finds all trajectories within a given distance of a query trajectory. We develop three GPU distance threshold search implementations that use indexing techniques significantly different from those used in CPU implementations. We determine experimentally under which conditions each approach performs well using one real-world astrophysics dataset and two synthetic datasets. Overall, we find that the GPU is an attractive technology for a broad range of relevant trajectory database scenarios.
January 19, 2015 by hgpu
