Techniques for Onboard Prioritization of Science Data for Transmission
Rebecca Castaño, Robert Charles Anderson, Tara A. Estlin, Dennis DeCoste, Forest Fisher, Daniel M. Gaines, Daniele Mazzoni, Michele Judd · 2003
Future planetary exploration missions will continue to face an ever-increasing data prioritization problem as instrument data collection rates continue to exceed spacecraft downlink transmission rates. Hard decisions must be made about what data are sent back to Earth and what data are purged without being seen by scientists. In this article, we present a suite of techniques for the prioritization of science data for transmission. These techniques include methods to ensure that data representative of the local geology, as well as any unusual observations, are given the highest priority. I. Introduction As planetary exploration continues to expand, the combination of more missions, an increased number of instruments, and the advanced capabilities of those instruments will cause an increase in the volume of data to be transmitted back to Earth via the Deep Space Network (DSN). Missions will have to make critical decisions regarding the quantity and quality of the downlinked data, both of which are indirectly affected by, among other things, the availability constraints of the DSN and the limited spacecraft power. Although the DSN’s receiving capability increases every year, the number of missions it must service is growing rapidly. New methods must be developed to maximize the science return for the available bandwidth. Conventional data compression helps in this regard; however, ever-increasing amounts of compression can cause unacceptably high distortion levels. We are developing onboard analysis methods to autonomously prioritize data for downlink as another approach to getting the most out of the limited bandwidth. Data prioritization already implicitly occurs. A priori decisions are made regarding when to turn on an instrument based on the number of data sets that can be transmitted to Earth. Instruments may not be activated because existing scenarios have limited or no ability to make intelligent decisions about the value of data being collected. In contrast, by collecting more data than can be transmitted to Earth and applying onboard data-analysis technology to carefully select which data are sent, the quality of data returned can be increased, maximizing the science return for the available downlink. Selecting the most