Efficient Algorithms for Historical Continuous kNN Query Processing over Moving Object Trajectories

Abstract

In this paper, we investigate the problem of efficiently processing historical continuous k-Nearest Neighbor (HCkNN) queries on R-treelike structures storing historical information about moving object trajectories. The existing approaches for HCkNN queries need high I/O (i.e., number of node accesses) and CPU costs since they follow depth-first fashion. Motivated by this observation, we present two algorithms, called HCP-kNN and HCT-kNN, which deal with the HCkNN retrieval with respect to the stationary query point and the moving query trajectory, respectively. The core of our solution employs best-first traversal paradigm and enables effective update strategies to maintain the nearest lists. Extensive performance studies with real and synthetic datasets show that the proposed algorithms outperform their competitors significantly in both efficiency and scalability.