Authors:
Hassan A. Karimi
and
Duangduen Roongpiboonsopit
Affiliation:
University of Pittsburgh, United States
Keyword(s):
Cloud computing, Geoprocessing, Real-time, Data-intensive, Geospatial data.
Related
Ontology
Subjects/Areas/Topics:
Cloud Application Architectures
;
Cloud Application Scalability and Availability
;
Cloud Computing
;
Cloud Computing Enabling Technology
;
Cloud Ilities (Scalability, Availability, Reliability)
;
Development Methods for Cloud Applications
;
Performance Development and Management
;
Platforms and Applications
Abstract:
Interest in implementing and deploying many existing and new applications on cloud platforms is continually growing. Of these, geospatial applications, whose operations are based on geospatial data and computation, are of particular interest because they typically involve very large geospatial data layers and specialized and complex computations. In general, problems in many geospatial applications, especially those with real-time response, are compute- and/or data-intensive, which is the reason why researchers often resort to high-performance computing platforms for efficient processing. However, compared to existing high-performance computing platforms, such as grids and supercomputers, cloud computing offers new and advanced features that can benefit geospatial problem solving and application implementation and deployment. In this paper, we present a distributed algorithm for geospatial data processing on clouds and discuss the results of our experimentation with an existing cloud
platform to evaluate its performance for real-time geoprocessing.
(More)