(MAPLD Conference)
PCD10:
"Making a Case for Distributed Adaptive Computing in Remote Sensing Science Data Processing"
A series of new remote sensing satellites will be launched by NASA beginning with the construction of the EOS-AM satellite to be launched in 1999. These satellites will harbor instruments that will collect scientific data from the earth producing much higher amounts of data than traditional satellites. Hence, there is a need for more advanced computing resources to process this data both onboard the satellite as well as on the ground.
Reconfigurable or adaptive computing represents a potential solution to remote sensing science data processing applications. These applications not only require massive processing resources, but also require large amounts of input/output data to be stored. Reconfigurable computing can provide application-specific customization that translates into performance improvement and a reduction in development time over custom application-specific integrated circuit implementation.
This poster presentation directly addresses the issues related to rapid prototyping of remote sensing science data processing applications using adaptive computers. Our results demonstrate that the use of the Java object-oriented programming language together with an adaptive computing platform significantly reduces development time, for a typical remote sensing application, while improving performance by an order of magnitude. An interesting result, obtained by executing a remote sensing science data processing application benchmark from a remote site demonstrates that a distributed adaptive computing implementation can be preferred over a traditional software implementation.
Home
Last Revised: July 03, 2002
Digital Engineering Institute
Web Grunt: Richard Katz
