ebook img

Pre-Hardware Optimization of Spacecraft Image Processing Algorithms and Hardware Implementation PDF

1 Pages·2002·0.08 MB·English
Save to my drive
Quick download
Download
Most books are stored in the elastic cloud where traffic is expensive. For this reason, we have a limit on daily download.

Preview Pre-Hardware Optimization of Spacecraft Image Processing Algorithms and Hardware Implementation

Pre-Hardware Optimization of Spacecraft Image Processing Software Algorithms and Hardware Implementation Semion Kizhner, David J. Petrick, Thomas P. Flatley, Phyllis Hestnes, Marit Jentoft-Nilsen National Aeronautics and Space Administration Abstract the PC environment, the necessity of RC hardware implementation for bottleneck code Spacecraft telemetry rates and telemetry product became more evident. The problem was solved complexity have steadily increased over the last beginning with the methodology described in decade presenting a problem for real-time [1], [2], and implementing a novel methodology processing by ground facilities. This paper for this specific application. The PC-RC proposes a solution to a related problem for the interface bandwidth problem for the class of Geostationary Operational Environmental applications with moderate input-output data Spacecraft (GOES-8) image data processing and rates but large intermediate multi-thread data color picture generation application. Although streams has been addressed and mitigated. This large super-computer facilities are the obvious opens a new class of satellite image processing heritage solution, they are very costly, making it applications for bottleneck problems solution imperative to seek a feasible alternative using RC technologies. The issue of a science engineering solution at a fraction of the cost. The algorithm level of abstraction necessary for RC proposed solution is based on a Personal hardware implementation is also described. Computer (PC) platform and synergy of Selected Matlab functions already implemented optimized software algorithms, and re- in hardware were investigated for their direct configurable computing hardware (RC) applicability to the GOES-8 application with the technologies, such as Field Programmable Gate intent to create a library of Matlab and IDL RC Arrays (FPGA) and Digital Signal Processors functions for ongoing work. A complete class of (DSP). It has been shown in [11 and [2] that this spacecraft image processing applications approach can provide superior inexpensive development using re-configurable computing performance for a chosen application on the technology to meet real-time requirements, ground station or on-board a spacecraft. including methodology, performance results and However, since RC technologies are still comparison with the existing system, is maturing, intensive pre-hardware steps are described in this paper. necessary to achieve the benefits of hardware implementation. This paper describes these steps Table of Context for the GOES-8 application - a software project originally developed using the Interactive Data Abstract Language (IDL) (Trademark of Research Introduction Systems, Inc.) on a Workstation/UNIX platform. 1.0 Problem engineering complexity and The solution involves converting the application solution methodology to a PC/Windows/RC platform, selected mainly 2.0 Heritage application and its operations by the availability of low cost, adaptable high- overview speed RC hardware. In order for the application 3.0 Porting the UNIX/IDL application to the to run on this PC/RC hybrid system, the IDL PC/Windows platform, modifying code, software was modified to account for platform compilation and run differences. It was interesting to examine the 4.0 Application performance timing gains and losses in performance on the new methodology platform, as well as unexpected observations 5.0 Timing tests on the PC/Windows98 platform before implementing RC hardware. After 6.0 Algorithm definition, optimization and RC substantial pre-hardware optimization steps in hardware implementation

See more

The list of books you might like

Most books are stored in the elastic cloud where traffic is expensive. For this reason, we have a limit on daily download.