Nearest Neighbor Queries in Shared-Nothing Environments

  • Authors:

  • Apostolos Papadopoulos;Yannis Manolopoulos

  • Affiliations:

  • Department of Informatics, Aristotle University, Thessaloniki, 54006 Greece/ apapadop@athena.auth.gr;Department of Informatics, Aristotle University, Thessaloniki, 54006 Greece/ manolopo@athena.auth.gr

  • Venue:
  • Geoinformatica
  • Year:
  • 1997

Quantified Score

Hi-index 0.00

Visualization

Abstract

In this paper, we propose an efficient solution to the problem ofnearest neighbor query processing in declustered spatial databases. Recentlya branch-and-bound nearest neighbor finding (BB-NNF) algorithm has beendesigned to process nearest neighbor queries in R-trees. However, thisalgorithm is strictly serial (branch-and-bound oriented) and its performancedegrades, during processing of a nearest neighbor query, if applied to aparallel environment, since it does not exploit any kind of parallelization.We develop an efficient query processing strategy for parallel nearestneighbor finding (P-NNF), assuming a shared nothing multi-processorarchitecture, where the processors communicate via a network. In our method,the relevant sites are activated simultaneously. In order to achieve thisgoal, statistical information is used. The efficiency measure is theresponse time of a given query. Experimental results, based on real-life andsynthetic datasets, show that the proposed method outperforms thebranch-and-bound method by factors.