Distributed k-NN query processing for location services

Abstract

The architecture named the OALIS is a cluster-based distributed computing system architecture which has been devised to efficiently handle a large volume of LBS application data. In this paper, we propose a distributed k-NN query processing scheme for moving objects on multiple computing nodes, each of which keeps records relevant to a different geographical zone. We also propose a hybrid k-NN scheme, which utilizes range queries instead of k-NN queries for the neighboring overlapped nodes, thus resulting in 30% reduction of query processing cost, Through some experiments, we show the efficiency of hybrid k-NN scheme over naïve k-NN scheme. © IFIP International Federation for Information Processing 2007.