proposed data-aware scheduling algorithm is much more 
efficient than the traditional FIFO method when a neighbor 
requirement is present in the user’s processing algorithm. 
6. CONCLUSION 
Parallel computing has been increasingly used to solve data- 
intensive problems in geospatial science. Inspired by these 
problems, this paper proposed a universal parallel framework 
for processing massive LiDAR point clouds in a HPC 
environment. Within this framework, the user/programmer is 
supported with a predefined Split-and-Merge programming 
paradigm. In this paradigm, user/programmers can focus on the 
simple functional expression of their specific algorithm into two 
distinct programs, Split and Merge, and leave parallelization and 
scheduling to the runtime system. This framework automatically 
and intelligently handles key scheduling decisions for tasks and 
data. For considering data sharing between task inputs, a 
specific data-aware scheduling algorithm is proposed to 
decrease the data communication time. One common LiDAR 
algorithm, DT, was evaluated to prove the efficiency and 
suitability of our proposed framework. 
ACKNOWLEDGEMENTS 
This work is supported by the Natural Science Foundation of 
China (Grant: 40971211 and 40721001). 
REFERENCES 
Borkar, S. and Chien, A., 2011. The future of microprocessors. 
Communications of the ACM, 54(5), pp.67-77. 
Dean, J. and Ghemawat, S., 2008. Mapreduce: Simplified data 
processing on large clusters. Communications of the ACM, 51(1), 
pp. 107-113. 
Guan, Q. and Clarke, K., 2010. A general-purpose parallel raster 
processing programming library test application using a 
geographic cellular automata model. International Journal of 
Geographical Information Science, 24(5), pp.695-722. 
Hawick, K., Coddington, P. and James, H., 2003. Distributed 
Frameworks and Parallel Algorithms for Processing Large-Scale 
Geographic Data. Parallel Computing, 29(10), pp.1297-1333. 
Jonker, P., Olk, J., and Nicolescu, C., 2008. Distributed bucket 
processing: A paradigm embedded in a framework for the 
parallel processing of pixel sets. Parallel Computing, 34(12), 
pp.735-746. 
Staples, G., 2006. TORQUE resource manager. In: Proceedings 
of the 2006 ACM/IEEE conference on Supercomputing. Tampa, 
Florida. 
Tehranian, S., Zhao, Y., Harvey, T., Swaroop, A. and Mckenzie, 
K., 2006. A robust framework for real-time distributed 
processing of satellite data. Journal of Parallel and Distributed 
Computing, 66(3), pp.403-418. 
Wang, H., Fu, X., et al., 2011. A common parallel computing 
framework for modeling hydrological processes of river basins. 
Parallel Computing, 37(6-7), pp. 302-315. 
Wu, H., Guan, X. and Gong, J., 2011. ParaStream: A parallel 
streaming Delaunay triangulation algorithm for LiDAR Points 
on Multicore Architectures. Computers & Geosciences, 37(9), 
pp.1355-1363. 
206 
KE 
tec] 
per 
rem 
ear 
tem 
for 
unc 
inte 
inf 
20€ 
inf 
fea 
pro 
Im: 
dev 
inte 
aut 
inte 
intc 
stai 
inte 
and 
bec 
of 
C 
Em