Search

One of the key research issues with moving objects databases is the uncertainty management. The uncertainty management for moving objects has been well studied recently, with many models and algorithms proposed. In this chapter, we analyze the uncertainty of moving objects in spatial networks and introduce an uncertain trajectory model and an index framework, the uncertain trajectory based Rtree (UTR-tree), for indexing the fully uncertain trajectories of network-constrained moving objects. Then, we introduce how to process queries on this framework. The content of this chapter is mainly from the work of Ding in [14]. [ABSTRACT FROM AUTHOR]

The trajectory prediction is an important part for the management of moving objects. For example, it can be used to improve the performance of the location update strategy and to support the predictive index and queries. In this chapter, we first review some linear prediction methods and analyze their problem in handling moving objects in spatial networks, and then present our simulation-based prediction methods: Fast-Slow Bounds Prediction and Time-Segment Prediction. [ABSTRACT FROM AUTHOR]

So far, we have introduced the basic querying for moving objects. There are still some advanced querying for moving objects. It is more difficult to deal with these queries. In this chapter, we introduce a few advanced queries, especially similar trajectory queries and density queries for moving objects. The goal of similar trajectory queries is to find the moving patterns in the trajectories of moving objects, while density queries are to efficiently find dense areas with high concentration of moving objects. We will discuss how to process both the snapshot and continuous density queries in this chapter. [ABSTRACT FROM AUTHOR]

Once we build the model and index for moving objects, we can answer the queries for moving objects. There are many types of queries in moving objects databases such as the nearest neighbor (NN) query, range query, and density query. In this chapter, we will introduce the basic querying types for moving objects according to spatial predicates, temporal predicates, and moving spaces. Though there are many techniques to support moving objects queries, most of the existing studies consider Euclidean spaces, where the distance between two objects is determined solely by their relative position in space. However, in practice, objects can usually move only on a pre-defined set of trajectories as specified by the underlying network. Hence we will introduce how to answer range queries and NN queries for moving objects in a spatial network, which is based on the work of Papadias in [11]. [ABSTRACT FROM AUTHOR]

For querying large amounts of moving objects, a key problem is to create efficient indexing structures that make it possible to effectively answer various types of queries. Traditional spatial indexing approaches cannot be used because the locations of moving objects are highly dynamic, which leads to frequent updates of index structures, which in turn will cause huge overheads. In this chapter, we first introduce a few of the underlying spatial index structures including the R-tree, Grid File, and Quad-tree. Then, we propose the indexing methods for moving objects in Euclidean space and in spatial networks. Finally, we describe techniques that index the past, present, and anticipated future positions of moving objects. [ABSTRACT FROM AUTHOR]

In moving objects applications, large numbers of locations can be sampled by sensors or GPS periodically, then sent from moving clients to the server and stored in a database. Therefore, continuously maintaining in a database the current locations of moving objects by using a tracking technique becomes very important. The key issue is minimizing the number of updates, while providing precise locations for query results. In this chapter, we will introduce some underlying location update methods. Then, we describe two location update strategies in detail, which can improve the performance. One is the proactive location update strategy, which predicts the movement of moving objects to lower the update frequency; the other is the group location update strategy, which groups the objects to minimize the total number of objects reporting their locations. [ABSTRACT FROM AUTHOR]

Location modeling is the foundation for moving objects databases. Existing database management systems are not well equipped to handle continuously changing data, such as the position of moving objects. The reason for this is that in traditional databases, data is assumed to be constant unless it is explicitly modified. This is unsatisfactory for MOD since locations of moving objects are continuously changing. In this chapter, we introduce a few underlying location modeling methods and propose a new graph of cellular automata (GCA) model to integrate the traffic movement features into the model of moving objects and the underlying spatial network. [ABSTRACT FROM AUTHOR]