A predictive spatio-temporal query retrieves the set of moving objects that will intersect a query window during a future time interval. Currently, the only access method for processing such queries in practice is the TPR-tree. In this paper we first perform an analysis to determine the factors that affect the performance of predictive queries and show that several of these factors are not considered by the TPR-tree, which uses the insertion/deletion algorithms of the R*-tree designed for static data. Motivated by this, we propose a new index structure called the TPR*- tree, which takes into account the unique features of dynamic objects through a set of improved construction algorithms. In addition, we provide cost models that determine the optimal performance achievable by any data-partition spatio-temporal access method. Using experimental comparison, we illustrate that the TPR*-tree is nearly-optimal and significantly outperforms the TPR-tree under all conditions.

predicted time 2 at positions predicted time 1 at

this paper we present probabilistic cost models that estimate the selectivity of spatio-temporal window queries and joins, and the expected distance between a query and its nearest neighbor(s). Our models capture any query/object mobility combination (moving queries, moving objects or both) and any data type (points and rectangles) in arbitrary dimensionality. In addition, we develop specialized spatio-temporal histograms, which take into account both location and velocity information, and can be incrementally maintained. Extensive performance evaluation verifies that the proposed techniques produce highly accurate estimation on both uniform and non-uniform data

Abstract not found

Abstract — Moving object databases have received considerable attention recently. Previous work has concentrated mainly on modeling and indexing problems, as well as query selectivity estimation. Here we introduce a novel problem, that of addressing density-based queries in the spatio-temporal domain. For example: “Find all regions that will contain more than 500 objects, ten minutes from now”. The user may also be interested in finding the time period (interval) that the query answer remains valid. We formally define a new class of density-based queries and give approximate, on-line techniques that answer them efficiently. Typically the threshold above which a region is considered to be dense is part of the query. The difficulty of the problem lies in the fact that the spatial and temporal predicates are not specified by the query. The techniques we introduce find all candidate dense regions at any time in the future. To make them more scalable we subdivide the spatial universe using a grid and limit queries within a pre-specified time horizon. Finally, we validate our approaches with a thorough experimental evaluation. I.

Abstract — Many novel spatio-temporal applications deal with moving objects. In such environments, a database typically maintains the initial position and the moving function for each object. Instead of updating the database whenever an object position changes (which is not manageable), updates are issued whenever the moving function deviates beyond a given threshold. For simplicity, we assume that objects move with linear trajectories. Maintaining the moving functions in a database introduces novel problems. For example, the database can answer queries about object positions in the future: “find all objects that will be in area A, 10 minutes from now”. In this paper we present a thorough performance evaluation of techniques for estimating the selectivity of such queries. We consider various existing estimators that can be stored in main memory and are updated dynamically. Furthermore, we propose two new approaches, a technique that uses histograms and a secondary index based estimator. We run a diverse set of experiments to identify the strengths and weaknesses of every approach, using a wide variety of datasets. I.

Moving objects (e.g., vehicles in road networks) continuously generate large amounts of spatio-temporal information in the form of data streams. Efficient management of such streams is a challenging goal due to the highly dynamic nature of the data and the need for fast, on-line computations. In this paper we present a novel approach for approximate query processing about the present, past, or the future in spatio-temporal databases. In particular, we first propose an incrementally updateable, multi-dimensional histogram for present-time queries. Second, we develop a general architecture for maintaining and querying historical data. Third, we implement a stochastic approach for predicting the results of queries that refer to the future. Finally, we experimentally prove the effectiveness and efficiency of our techniques using a realistic simulation. 1.

Spatio-temporal databases store information about the positions of individual objects over time. However, in many applications such as traffic supervision or mobile communication systems, only summarized data, like the number of cars in an area for a specific period, or phone-calls serviced by a cell each day, is required. Although this information can be obtained from operational databases, its computation is expensive, rendering online processing inapplicable. In this paper, we present specialized methods, which integrate spatio-temporal indexing with pre-aggregation. The methods support dynamic spatio-temporal dimensions for the efficient processing of historical aggregate queries without a priori knowledge of grouping hierarchies. The superiority of the proposed techniques over existing methods is demonstrated through a comprehensive probabilistic analysis and an extensive experimental evaluation.

Dealing with moving objects necessitates having available complete geographical traces for determining exact or possible locations that objects have had, have or will have. This is where trajectory determination plays an important role, and on which classification, aggregation and comparison methods must be built. The purpose of aggregation is to identify similar trajectories and to represent them by a single trajectory. Although much work has been done in similarity measurements for time series data, they mainly deal with one dimensional time series. On the other hand, they are good for short time series and in absence of noise, which is definitely not the case for moving objects. This paper describes di#erent approaches to aggregate similar trajectories.

Given a region qR and a future timestamp qT, a “range aggregate ” query estimates the number of objects expected to appear in qR at time qT. Currently the only methods for processing such queries are based on spatiotemporal histograms, which have several serious problems. First, they consume considerable space in order to provide accurate estimation. Second, they incur high evaluation cost. Third, their efficiency continuously deteriorates with time. Fourth, their maintenance requires significant update overhead. Motivated by this, we develop Venn sampling (VS), a novel estimation method optimized for a set of “pivot queries ” that reflect the distribution of actual ones. In particular, given m pivot queries, VS achieves perfect estimation with only O(m) samples, as opposed to O(2 m) required by the current state of the art in workload-aware sampling. Compared with histograms, our technique is much more accurate (given the same space), produces estimates with negligible cost, and does not deteriorate with time. Furthermore, it permits the development of a novel “query-driven ” update policy, which reduces the update cost of conventional policies significantly. 1.