Date of Award
3-23-2018
Document Type
Thesis
Degree Name
Master of Science in Electrical Engineering
Department
Department of Electrical and Computer Engineering
First Advisor
Dane F. Fuller, PhD.
Abstract
Simplified General Perturbations Number 4 (SGP4) has been the traditional algorithm for performing Orbit Determination (OD) onboard orbiting spacecraft. However, the recent rise of high-performance computers with low Size, Weight, and Power (SWAP) factors has provided the opportunity to use Special Perturbations (SP), a more accurate algorithm to perform onboard OD. This research evaluates the most efficient way to implement SP on NVIDIA’s Jetson TX series of integrated Graphical Processing Units (GPUs). An initial serial version was implemented on the Jetson TX1 and TX2's Central Processing Units (CPUs). The runtimes of the initial version are the benchmark that the runtimes of the other versions were compared against. A second version of SP was implemented using compiler optimizations to increase the speed of the program. A third version was developed to utilize the Jetsons' 256-core GPU for parallel processing to reduce the runtimes of the program. Runtimes of the different versions were then analyzed to determine the most efficient way to implement SP on the Jetson TX series of computers.
AFIT Designator
AFIT-ENG-MS-18-M-047
DTIC Accession Number
AD1056191
Recommended Citation
Moore, Tyler M., "Special Perturbations on the Jetson TX1 and TX2 Computers" (2018). Theses and Dissertations. 1816.
https://scholar.afit.edu/etd/1816