| Summary: | The Air Force Space Command (AFSPACECOM) uses an analytic satellite motion model based an the Brouwer theory, referred to as SGP4, for near-earth objects, and another satellite motion model based on the work of Hujsak, referred to as SDP4, for deep-space objects. These models assist in tracking over 7000 objects around the Earth each day. original satellite motion model used by the AFSPACECOM based on the work of Kozai and Brouwer, known as the Simplified General Perturbations model or SGP, is an analytic model still being used at several sites. Based on the increasing number of space objects that require tracking and the desire for accuracy, there is a growing need to reduce computation time in implementing satellite motion models. Parallel computing offers one method to achieve this objective. This thesis investigates the parallel computing potential of SGP, SGP4, and SDP4 using the Intel iPSC/2 hypercube multicomputer. This thesis chooses a parallel algorithm and applies it to SGP, SGP4, and SDP4 for implementation on the hypercube, and reports on the potential reduction in computer time by first considering only the calculational portion of the algorithm and then including the effects of the reading and writing portions. A diskette containing the Fortran software developed is available upon request from bneta moon.math.nps,navy.mil
|
|---|
