OPTIMIZATION APPROACHES TO MPI AND AREA MERGING-BASED PARALLEL BUFFER ALGORITHM

• Autor(a) principal: FAN, JUNFU
• Data de Publicação: 2014
• Outros Autores: JI, MIN, GU, GUOMIN, SUN, YONG
• Tipo de documento: Artigo
• Idioma: por
• Título da fonte: Boletim de Ciências Geodésicas
• Texto Completo: https://revistas.ufpr.br/bcg/article/view/36797
• topic: Geociências; Geodésia; Area Merging; Parallel Buffer; Task Partition; Vertex Accumulation Effect; Tree-like Merging; MPI.

On buffer zone construction, the rasterization-based dilation method inevitablyintroduces errors, and the double-sided parallel line method involves a series ofcomplex operations. In this paper, we proposed a parallel buffer algorithm based onarea merging and MPI (Message Passing Interface) to improve the performances ofbuffer analyses on processing large datasets. Experimental results reveal that thereare three major performance bottlenecks which significantly impact the serial andparallel buffer construction efficiencies, including the area merging strategy, thetask load balance method and the MPI inter-process results merging strategy.Corresponding optimization approaches involving tree-like area merging strategy, the vertex number oriented parallel task partition method and the inter-processresults merging strategy were suggested to overcome these bottlenecks. Experimentswere carried out to examine the performance efficiency of the optimized parallelalgorithm. The estimation results suggested that the optimization approaches couldprovide high performance and processing ability for buffer construction in a clusterparallel environment. Our method could provide insights into the parallelization ofspatial analysis algorithm.