IN13D-07:
High-Speed on-Board Data Processing for Science Instruments

Monday, 15 December 2014: 3:10 PM
Jeffrey Beyon, Tak-Kwong Ng, Mitchell Jordan Davis and Bing Lin, NASA Langley Research Center, Hampton, VA, United States
Abstract:
A new development of on-board data processing platform has been in progress at NASA Langley Research Center since April, 2012, and the overall review of such work is presented. The project is called High-Speed OnBoard Data Processing for Science Instruments (HOPS) and focuses on an air/space-borne high-speed scalable data processing platform for three particular National Research Council’s Decadal Survey missions such as Active Sensing of CO2 Emissions over Nights, Days, and Seasons (ASCENDS), Aerosol-Cloud-Ecosystems (ACE), and Doppler Aerosol Wind Lidar (DAWN) 3-D Winds. HOPS utilizes advanced general purpose computing with Field Programmable Gate Array (FPGA) based algorithm implementation techniques. The significance of HOPS is to enable high speed on-board data processing for current and future science missions with its reconfigurable and scalable data processing platform. A single HOPS processing board is expected to provide approximately 66 times faster data processing speed for ASCENDS, more than 70% reduction in both power and weight, and about two orders of cost reduction compared to the state-of-the-art (SOA) on-board data processing system. Such benchmark predictions are based on the data when HOPS was originally proposed in August, 2011. The details of these improvement measures are also presented. The two facets of HOPS development are identifying the most computationally intensive algorithm segments of each mission and implementing them in a FPGA-based data processing board. A general introduction of such facets is also the purpose of this presentation.