Efficient processing of distance–time k th‐order skyline queries in bicriteria networks

Abstract

With the increasing complexity of traffic conditions in road networks, the nearest destination cannot be necessarily reached in the fastest time. The traditional nearest neighbor (NN) and kNN searches in spatial network databases with single cost criterion are often strongly restrictive. In this paper, the authors consider the problem of kth-order skyline queries in bicriteria networks, where edges represent road segments. Their proposed kth-order skyline queries consider distance, time preferences of each edge, thus having two kinds of skyline queries, named distance optimal kth-order skyline query (DO-kOSQ) and time optimal kth-OSQ (TO-kOSQ). They design algorithms for the two kinds of skyline queries in bicriteria networks based on incremental network expansion method and further develop a maximum distance/time restriction strategy to improve the efficiency of the algorithms. Experimental results show that all of their methods are far below 1000 input–output input/output (IO) accesses and 1 s of central processing unit (CPU) time. For real road networks, their kOSQ+ queries need only 51.6% IO accesses and 59.6% CPU time of those for kOSQ queries, whereas for the larger road network the percentages are 51.8% and 51.2%, respectively. Thus, the results indicate the efficiency and effectiveness of their proposed methods.