Citation Context

...nd a small 3D Rtree for time-interval queries. This indexing structure has been shown to outperform other historical trajectory indexing structures, such as the popular TB-tree [23]. SEB[32] and SETI =-=[8]-=- are historical trajectory indexing techniques that partition the spatial extents, and build indices on the temporal dimension. A number of indexing methods have also focused on efficient methods for ...

Citation Context

... left to right. The left-most leaf node is the first inserted node and the right-most leaf node is the last inserted one. The TB-tree is an extension of the STR-tree to handle only trajectories. SETI =-=[8]-=-: The Scalable and Efficient Trajectory Index (SETI) partitions the spatial dimension into static, non-overlapping partitions. The main observation is that the change of the spatial dimension is limit...

Citation Context

...o space limitations. To compute the boundary-node distance estimator (1) we partition the space into non-overlaping cells. Nonoverlaping space partinioning has appeared before in the literature, e.g. =-=[2]-=-. A boundary node [9] of a cell is a node directly connected with some other node in a different cell. That is, any path linking a node in a cell C1 with some node in a different cell C2 must go throu...

Citation Context

...dicates, and temporal constraints. Any trajectory indexing scheme that can answer nearest neighbor, range queries, and other spatial predicates can be used with the STP query algorithm described here =-=[22, 15, 2, 13, 18, 21]-=-. were there is an inherent uncertainty in moving object trajectories [23]. 6 Conclusions We introduced and formalized a novel type of spatiotemporal pattern query for trajectories. We designed specia...

Citation Context

...ing objects, a number of R-tree-based methods for indexing moving objects have been proposed. They focus on one of the following approaches: (1) Indexing the historical trajectories of objects, e.g., =-=[3, 7, 12, 13, 22, 23, 25]-=-; (2) Indexing the current locations of objects, e.g., [5, 9, 14, 15, 18, 19]; and (3) Indexing the predicted trajectories of objects, e.g., [16, 20, 24]. Most of these works assume that the updates a...

Citation Context

...ies. In the context of spatio-temporal databases, several spatio-temporal access methods [22] are proposed that are scalable in terms of the number of moving objects that can be supported (e.g., see =-=[5, 24, 25, 27, 28, 32, 34]-=-). The scalability in terms of the number of spatio-temporal queries and moving objects is addressed in [25]. In addition, each spatiotemporal access method is concerned with a specific type of spatio...

Citation Context

...io-temporal Databases Existing algorithms for continuous spatio-temporal query processing focus mainly on materializing incoming spatio-temporal data in disk-based index structures (e.g., hash tables =-=[12,49]-=-, grid files [21,39,44], the Btree [29], the R-tree [33,35], and the TPR-tree [48,52]). Thus, it is implicitly assumed that all incoming data can be stored. Scalable execution of continuous spatiotemp...

Citation Context

... in this work since movement is representing as volume data for discrete spatial locations. Further, various approaches to store and query trajectory data have been proposed in literature, e.g., [11] =-=[14]-=- [20]. However their distinct and decisive disadvantage is that they cannot easily be implemented in practice, i.e., by using and extending off-the-shelf DBMSes to manage the data. We are not aware of...