Efficient Continuous Nearest Neighbor Query in Spatial Networks Using Euclidean Restriction

Abstract

In this paper, we propose an efficient method to answer continuous k nearest neighbor (CkNN) queries in spatial networks. Assuming a moving query object and a set of data objects that make frequent and arbitrary moves on a spatial network with dynamically changing edge weights, CkNN continuously monitors the nearest (in network distance) neighboring objects to the query. Previous CkNN methods are inefficient and, hence, fail to scale in large networks with numerous data objects because: 1) they heavily rely on Dijkstra-based blind expansion for network distance computation that incurs excessively redundant cost particularly in large networks, and 2) they blindly map all object location updates to the network disregarding whether the updates are relevant to the CkNN query result. With our method, termed ER-CkNN (short for Euclidian Restriction based CkNN), we utilize ER to address both of these shortcomings. Specifically, with ER we enable 1) guided search (rather than blind expansion) for efficient network distance calculation, and 2) localized mapping (rather than blind mapping) to avoid the intolerable cost of redundant object location mapping. We demonstrate the efficiency of ER-CkNN via extensive experimental evaluations with real world datasets consisting of a variety of large spatial networks with numerous moving objects.

This research has been funded in part by NSF grants IIS-0238560 (PECASE), IIS-0534761,IIS-0742811 and CNS-0831505 (CyberTrust), and in part from CENS and METRANS Transportation Center, under grants from USDOT and Caltrans.Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation.