RT Journal Article
T1 A data relocation approach for terrain surface analysis on multi-GPU systems: a case study on the total viewshed problem.
A1 Sánchez Fernández, Andrés Jesús
A1 Romero-Gómez, Luis Felipe
A1 Bandera-Burgueño, Gerardo
A1 Tabik, Siham
K1 Sistemas de información geográfica
K1 Geografía - Proceso de datos
AB Digital Elevation Models (DEMs) are important datasets for modelling line-of-sightphenomena such as radio signals, sound waves and human vision. These are commonly analysed using rotational sweep algorithms. However, such algorithms requirelarge numbers of memory accesses to 2D arrays which, despite being regular, resultin poor data locality in memory. This paper proposes a new methodology calledskewed Digital Elevation Model (sDEM), which substantially improves the localityof memory accesses and largely increases the inherent parallelism involved in thecomputation of rotational sweep-based algorithms. In particular, sDEM applies adata restructuring technique before accessing the memory and performing the computation. To demonstrate the high efficiency of sDEM, we use the problem of totalviewshed computation as a case study. Different implementations for single-core,multi-core, single-GPU and multi-GPU platforms are provided. We conducted twoexperiments in which sDEM is compared with (i) the most used geographic information systems (GIS) software and (ii) the state-of-the-art algorithm. In the firstexperiment, sDEM is in average 8.8x faster than current GIS software despite beingable to consider only few points because of their limitations. In the second experiment, sDEM is 827.3x faster than the state-of-the-art algorithm, considering thebest studied case
PB Taylor & Francis
YR 2020
FD 2020-12-04
LK https://hdl.handle.net/10630/30394
UL https://hdl.handle.net/10630/30394
LA eng
NO International Journal of Geographical Information Science, Vol. 35, N. 8
NO Política de acceso abierto tomada de: https://v2.sherpa.ac.uk/id/publication/5400?template=romeo
DS RIUMA. Repositorio Institucional de la Universidad de Málaga
RD 24 ene 2026