Continuous visible k nearest neighbor query on moving objects

Abstract

A visible k nearest neighbor (VkNN) query retrieves k objects that are visible and nearest to the query object, where "visible" means that there is no obstacle between an object and the query object. Existing studies on the VkNN query have focused on static data objects. In this paper we investigate how to process the query on moving objects continuously. We propose an effective filtering-andrefinement framework for evaluating this type of queries. We exploit spatial proximity and visibility properties between the query object and data objects to prune search space under this framework. A detailed cost analysis and a comprehensive experimental study are conducted on the proposed framework. The results validate the effectiveness of the pruning techniques and verify the efficiency of the proposed framework. The proposed framework outperforms a straightforward solution by an order of magnitude in terms of both communication and computation costs.

Finding Visible kNN Objects in the Presence of Obstacles within the User’s View Field †

“…Numerous algorithms and data structures (e.g., B-tree [26], grid structure [27], or R-tree [28]) have been proposed for kNN query processing. Only a few of them are capable of processing "visible" kNN queries [4,14,15,[18][19][20][21][22][23][24][25]. We can classify these works into two types: (1) visible kNN queries with the obstacle constraint and (2) visible kNN queries with the view field constraint.…”

“…Gao et al also proposed several pruning heuristics to improve the query performance. Another work that focused on continuously determining visible k nearest neighbors was proposed by Wang et al [20]. They proposed a filtering-and-refinement framework to process the query efficiently.…”

“…Given a query q(l, r, θ), a set of data objects In addition, to represent the distance between two objects, taking visibility into consideration, we define the visible distance V d as follows: Definition 3. Visible distance [20] Given a query q and a data object p, if p is visible to q, the visible distance between q and p (denoted as V d (q, p)) is the minimum Euclidean distance between them (i.e., dist(q, p)). Otherwise,…”

“…Both real and synthetic datasets were used in our experiments. To reflect the real-world situations of our problem with the existence of obstacles, we used a real dataset that is widely used in the research of visible nearest neighbor queries [4,18,20,30], containing 24,650 river distribution information data in Greece [36] as our obstacle sets, and each of them is represented by a bounding rectangle. We took the diagonals of the rectangles as the obstacles in our experiments, and the distribution is shown in Figure 22.…”

Continuous Visible Query for Three-Dimensional Objects in Spatial Databases

“…Gu et al [15] proposed a Continuous k Nearest Neighbor (CkNN) query based on control point to improve CPU efficiency and I/O. Wang et al [16] achieved Visible k Nearest Neighbor (VkNN) query on moving objects by using periodic queries and -tree indexes.…”