GIS Task Parallelization on Computational Grid

초록

There is a growing demand of computational grid for next generation of GIS which requires high performance computation, operation of distributed multi-resources, and collaborative platforms for multiple end users. Computational grid with GIS integrates widely used parallel and distributed technologies and presents a high performance computing for solving data-intensive and computation-intensive GIS problems, and supports a collaborative GIS to sat-isfy QoS for end users. This paper focuses on the distribution and paralleliza-tion of a computation-intensive GIS problem on a computational grid. This pa-per reviews existing computational grid-based GIS applications and suggests a computational grid-based approach to perform GIS task partitioning and archi-val. The approach allows executing a complex computation to create compli-cated maps and to represent geo-spatial information on collaborative platforms. For performance evaluation, this paper develops a DEVS (Discrete Event Sys-tem specification) formalism-based GIS model with computational grid envi-ronment. Analytical and empirical results show reduction of system execution time using distribution and parallelization on a computational grid.