|
Article information
2017 , Volume 22, Special issue, p.60-74
Potapov V.P., Popov S.E., Kostylev M.A.
Method of Sar data processing based on massively parallel Apache Spark system
The paper presents a novel approach to development of distributed software for streaming pre- and post-processing of radar images. A distinctive feature of the system is employment of Apache Spark mass-parallel processing, which allowed operations with general-purpose hardware as well as the ability to execute existing algorithms that are not designed for distributed processing on multiple nodes without changing the implementation of the algorithm. The paper compares distributed computing technologies and presents a general description of the cluster and the mechanism for performing the tasks of pre- and post-processing of radar data, as well as specifies the implementation of tasks within the framework of the proposed approach. The parallel computations methods based on GPUs, Java-multithreading and the application development based on a massively parallel architecture are used. In particular, in the phase unwrapping algorithm, the construction of the growth ring, the estimating of absolute phase values and the performance of reliability tests performed on the side of the GPU are presented. The number map of the regions which is monitored and updated after each iteration allows monitoring all regions along the optimal unwrapping path simultaneously and tracking the moments of their encounters and intersections. In the massively parallel algorithm execution, we apply a pairing deployment scheme for the input data in separate HDFS directories. As result of reading of the input data, a Resilient Distributed Datasets object is created. It, consists of pairs which include the file name and the binary data stream. The binary sequence is passed to the map-function, which executes the phase unwrapping procedure. The implementation of the map-function checks the node hardware and runs the code on the CPU or GPU. The interferometric image is processed entirely locally at the computational node without dividing into parts. Parallelization is achieved by processing a large number of images on a set of nodes. In conclusion, the test results of cluster execution of the developed algorithms are presented. 7õ speedup performance is shown with parallel start of eight computational tasks in the phase unwrapping algorithm. Compared to traditional approaches to radar data processing, in which parallel computations either do not apply, or are used only to improve the performance of calculations, the proposed solution is focused on processing a large number of packet data.
[full text]
Keywords: Apache Spark, Apache Hadoop, distributed information systems, sar interfometry, processing algorithms
Author(s):
Potapov Vadim Petrovich
Dr. , Professor
Position: Deputy director
Office: Federal Research Center for Information and Computational Technologies
Address: 650003, Russia, Kemerovo
Phone Office: (3842) 211400
E-mail: potapov@ict.sbras.ru
SPIN-code: 8947-1880Popov Semen Evgenievich
PhD.
Position: Senior Research Scientist
Office: Federal Research Center for Information and Computational Technologies
Address: 630090, Russia, Novosibirsk, Lavrentiev avenue, 6
Phone Office: (905)9692107
E-mail: popov@ict.sbras.ru
SPIN-code: 5627-9584Kostylev Mikhail Alexandrovich
Position: Student
Office: Institute of Computational Technologies SO RAN
Address: 630090, Russia, Novosibirsk, Lavrentiev avenue, 6
Phone Office: (3842) 211400
References:
Ñïèñîê ëèòåðàòóðû / References
[1] Elizavetin I.V., Shuvalov R.I., Bush V.A. Principles and methods of SAR Interferometry for the purpose of forming a digital elevation model. Geodeziya i kartografiya. 2009; (1):39-45. (In Russ.)
[2] Ferretti A., Monti-Guarnieri A., Prati C., et al. InSAR Principles: Guidelines for SAR Interferometry Processing and Interpretation. Available at: http://www.esa.int/esapub/tm/tm19/TM-19_ptA.pdf (accessed 02.08.2016)
[3] Zhengxiao Li, James Bethel. Image coregistration in sar interferometry. The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences. Part B1. Beijing 2008; (XXXVII):433-438.
[4] Massonnet, D., Feigl, K. L. Radar interferometry and its application to changes in the earth’s surface. Reviews of Geophysics. 1998; 36(4):441-500.
[5] Costantini, M.; Farina, A.; Zirilli, F. A fast phase unwrapping algorithm for SAR interferometry. IEEE Transactions on Geoscience and Remote Sensing. 2002; 37(1):452-460.
[6] Perhaad Mistry, Sherman Braganza, David Kaeli, Miriam Leeser Accelerating phase unwrapping and affine transformations for optical quadrature microscopy using CUDA. Proceedings of 2nd Workshop on General Purpose Processing on Graphics Processing Units, GPGPU 2009. USA, Washington, DC: ACM; 2009: 28-37.
7] Karasev, P.A., Campbell, D.P., Richards, M.A. Obtaining a 35x Speedup in 2D Phase Unwrapping Using Commodity Graphics Processors. Radar Conference, 2007: 574-578.
[8] Verba, V.S., Neronskiy, L.B., Osipov, I.G., Turuk, V.E. Radiolokatsionnye sistemy zemle-obzora kosmicheskogo bazirovaniya [Radar systems of ground-based space survey]. Moscow: Radiotekhnika; 2010: 675. ( In Russ.)
[9] Wu, Z., Ma, W., Long, G., Li, Y., Tang, Q., Wang, Z. High Performance Two-Dimensional Phase Unwrapping on GPUs. Proceedings of the 11th ACM Conference on Computing Frontiers - CF ’14. New York, USA: ACM; 2014:35:1-35:10
[10] Shi Xin-Liang, Xie Xiao-Chun. GPU acceleration of range alignment based on minimum entropy criterion. Radar Conference 2013, IET International. 14-16 April 2013:1-4.
[11] Guerriero, A., Anelli, V. W., Pagliara, A., Nutricato, R., Nitti, D.O. High performance GPU implementation of InSAR time-consuming algorithm kernels. Proceedings of the 1st WORKSHOP on the State of the art and Challenges Of Research Efforts at POLIBA. Bari, Italy: Politecnico di Bari; 2014: 383.
[12] Zhang, F., Wang, B., Xiang, M. Accelerating InSAR raw data simulation on GPU using CUDA. Geoscience and Remote Sensing Symposium (IGARSS). 2010 IEEE International. 25-30 July 2010. Bari, Italy: Politecnico di Bari; 2010: 2932-2935.
[13] Marinkovic, P. S., Hanssen, R. F., Kampes, B. M. Utilization of Parallelization Algorithms in InSAR/PS-InSAR Processing. Proceedings of the 2004 Envisat ERS Symposium (ESA SP-572). 6-10 September 2004. Salzburg, Austria: ESA; 2004: 1-7.
[14] Sheng, G., Qi-Ming, Z., Jian, J., Cun-Ren, L., TONG, Q. Parallel processing of InSAR interferogram filtering with CUDA programming. Science of Surveying and Mapping. 2015; (1):54-68.
[15] Gabriel, E., Fagg, G. E., Bosilca, G., Angskun, Th., Dongarra, J.J., Squyres, J.M., Sahay, V., Kambadur, Pr., Barrett, B., Lumsdaine, A., Castain, R.H., Daniel, D.J., Graham, R.L., Woodall, T.S. Open MPI: Goals, Concept, and Design of a Next Generation MPI Implementation. Available at: https://www.open-mpi.org/papers/euro-pvmmpi-2004-overview/euro-pvmmpi-2004-overview.pdf (accessed 23.09.2014).
[16] Kampes, B., Hanssen, R., Perski, Z. Radar Interferometry with Public Domain Tools presentation. Available at: http://doris.tudelft.nl/Literature/kampes_fringe03.pdf (accessed 23.09.2014).
[17] Frigo, M., Johnson, S. G. FFTW: An Adaptive Software Architecture for the FFT. In ICASSP conference proceedings. 15 May 1998. Seattle, Washington, USA: IEEE; 1998; (3):1381- 1384.
[18] Larkin, J. Fast GPU Development with CUDA Libraries. Available at: https://www.olcf.ornl.gov/wp-content/uploads/2013/02/GPU_libraries-JL.pdf (accessed 23.09.2014).
[19] Demmel, J., Dongarra, J. ST-HEC: Reliable and Scalable Software for Linear Algebra Computations on High End Computers. Available at: https://people.eecs.berkeley.edu/~demmel/Sca-LAPACK-Proposal.pdf (accessed 23.09.2014).
[20] Feoktistov A. A., Zaharov A. I., Gusev M. A., Denisov, P.V. Investigation of fabilities f the method of small baselines technique on the example of SBaS module of SARScape software package and the data of the RSA ASAR / ENVISat and PALSAR / ALOS. Part 1. Key points of the method. Zhurnal Radioelektroniki . 2015; (9):26. (In Russ.)
[21] Zinno, I. , Mossucca, L., Elefante, S., De Luca, C., Casola, V., Terzo, O., Casu, F., Lanari, R. Cloud Computing for Earth Surface Deformation Analysis via Spaceborne Radar Imaging: a Case Study. IEEE Trans. Cloud Computing. 2016; (4):104-118.
[22] Zinno, I., Mossucca, L., Elefante, S., Mossucca, L., De Luca, C., Manunta, M., Terzo, O., Lanari, R., Casu, F. First Assessment of the P-SBAS DInSAR Algorithm Performances Within a Cloud Computing Environment. Journal of selected topics in applied earth observations and remote sensing. 2015; (8):4675 – 4686.
[23] Kannan, P. Beyond Hadoop MapReduce Apache Tez and Apache Spark. Available at: http://www.sjsu.edu/people/robert.chun/courses/CS259Fall2013/s3/F.pdf (accessed 02.08.2016)
[24] Nathan P. Real-Time Analytics with Spark Streaming. Available at: http://viva-lab.ece.virginia.edu/foswiki/pub/InSAR/RitaEducation/InSAR%20Technology%20Literature%20Search.pdf (accessed 02.08.2016)
[25] Nagler, E. Introduction to Oozie. // Apache Oozie Documentation. Available at: http://www.cse.buffalo.edu/~bina/cse487/fall2011/Oozie.pdf (accessed 02.08.2016)
[26] Jhajj, R. Apache Hadoop Hue Tutorial. Available at: https://examples.javacodegeeks.com/enterprise-java/apache-hadoop/apache-hadoop-hue-tutorial/ (accessed 02.08.2016)
Bibliography link:
Potapov V.P., Popov S.E., Kostylev M.A. Method of Sar data processing based on massively parallel Apache Spark system // Computational technologies. 2017. V. 22. XVII All-Russian Conference of Young Scientists on Mathematical Modeling and Information Technology. P. 60-74
|
