A classification scheme for spatial problems is presented and used to develop strategies for decomposing a popular shortest path algorithm into parallel processes. Several decompositions are implemented on a four-node Transputer Array, running in a host PC. Representative results are presented for shortest path problems with a range of network sizes and characteristics.
Conference proceeding
Parallel processing for network analysis: decomposing shortest path algorithms on MIMD computers
Proceedings 5th International Symposium on Spatial Data Handling, pp.682-691
Charleston, SC, USA
1992
Abstract
Details
- Title: Subtitle
- Parallel processing for network analysis: decomposing shortest path algorithms on MIMD computers
- Creators
- Marc P Armstrong - University of IowaPaul J Densham - State University of New York at BuffaloYuemin Ding - State University of New York at Buffalo
- Resource Type
- Conference proceeding
- Publication Details
- Proceedings 5th International Symposium on Spatial Data Handling, pp.682-691
- Conference
- Charleston, SC, USA
- Copyright
- Copyright © 1992 the authors
- Language
- English
- Date published
- 1992
- Academic Unit
- Geographical and Sustainability Sciences
- Record Identifier
- 9983557340802771
Metrics
113 File views/ downloads
43 Record Views