Scientific Data Processing on Mobile Devices

Transcription

1 Scientific Data Processing on Mobile Devices David Mascharka PI: Victor Pankratius MIT Haystack Observatory 6 August 2015

2 Smartphones for Scientific Data?

3 Smartphones for Scientific Data? Powerful multi-core computers with many sensors

4 Smartphones for Scientific Data? Powerful multi-core computers with many sensors Nearly ubiquitous [1]

5 What Mobile Devices Offer

6 What Mobile Devices Offer Ubiquity allows global distributed system Global sensor network

7 What Mobile Devices Offer Ubiquity allows global distributed system Global sensor network Data transfer from the field to scientists

8 What Mobile Devices Offer Ubiquity allows global distributed system Global sensor network Data transfer from the field to scientists Pre-processing, visualization, and validation in the field

9 What Mobile Devices Offer Ubiquity allows global distributed system Global sensor network Data transfer from the field to scientists Pre-processing, visualization, and validation in the field This research: configurable mobile pipeline for scientific data

10 Mahali Ionospheric total electron content (TEC) Global sensor network for monitoring TEC

11 Ionospheric Total Electron Content Insight into solar and terrestrial processes Coronal mass ejections Solar winds Earthquakes [2] Nuclear detonations [3] Atmospheric coupling - everything [4]

12 [5]

13 RINEX Files Collect Data Parsing Compute Phase Difference Get Satellite Time Series Processing Compute Differential Range Compute Phase Shift Remove Satellite Bias Calculate and Remove Receiver Bias Apply Mapping Function Compute Elevation Angle Find Closest Ephemeris Visualization Upload Data

14 Data Phase = L1-L2 Difference of 2 pseudorange values

15 Data Phase = L1-L2 Difference of 2 pseudorange values

16 Data Phase = L1-L2 Difference of 2 pseudorange values Shift phase to differential range Compute satellite position Calculate angle of elevation Apply mapping function

17 Data Collected outside Haystack using CASES receiver 1,018,799 points at 93,575 times Data collected every second for 26 hours starting midnight May 18

18 Devices Device Clock (GHz) # Cores RAM (GB) Android Build Runtime Moto X ART Bionic Dalvik Razr i ART Nexus ART Nexus ART W N/A OpenJDK 8

19 Results - May 18 (1 million points) Device Runtime (s) Memory Freed (MB) Pause Time (s) Moto X Bionic Razr i Nexus Nexus W

20

21 Data Validation Compared against Bill Rideout s Python TEC module [6] Data collected by UNAVCO in Atka, Alaska

22 Conclusions Mobile devices feasible and useful for scientific data Reusable pipeline for scientific data transfer, processing, visualization

23 Acknowledgements Mahali funded by NSF INSPIRE AGS (PI: V. Pankratius) REU program supported by the NSF AST Collaborators at Boston College, Virginia Tech, Johns Hopkins University, Colorado State University UNAVCO for loans of dual-frequency GPS receivers Intel for loans of smartphones Haystack

24 References I D. Gilstrap, Ericsson mobility report on the pulse of the networked society, G. Occhipinti, L. Rolland, P. Lognonné, and S. Watada, From sumatra 2004 to tohoku-oki 2011: the systematic gps detection of the ionospheric signature induced by tsunamigenic earthquakes, Journal of Geophysical Research: Space Physics, vol. 118, no. 6, pp , J. Park, R. R. von Frese, D. A. Grejner-Brzezinska, Y. Morton, and L. R. Gaya-Pique, Ionospheric detection of the 25 may 2009 north korean underground nuclear test, Geophysical Research Letters, vol. 38, no. 22, 2011.

25 References II J. L. Chau, L. P. Goncharenko, B. G. Fejer, and H.-L. Liu, Equatorial and low latitude ionospheric effects during sudden stratospheric warming events, Space science reviews, vol. 168, no. 1-4, pp , V. Pankratius, F. Lind, A. Coster, P. Erickson, and J. Semeter, Mobile crowd sensing in space weather monitoring: the Mahali project, Communications Magazine, IEEE, vol. 52, pp , Aug W. Rideout and A. Coster, Automated gps processing for global total electron content data, GPS Solutions, vol. 10, no. 3, pp , 2006.

26 Questions