Wireless sensor networks are proving to be useful in a variety of settings. A core challenge in these networks is to minimize energy consumption. Prior database research has proposed to achieve this by pushing data-reducing operators like aggregation and selection down into the network. This approach has proven unpopular with early adopters of sensor network technology, who typically want to extract complete “dumps ” of the sensor readings, i.e., to run “SELECT *” queries. Unfortunately, because these queries do no data reduction, they consume significant energy in current sensornet query processors. In this paper we attack the “SELECT * ” problem for sensor networks. We propose a robust approximate technique called Ken that uses replicated dynamic probabilistic models to minimize communication from sensor nodes to the network’s PC base station. In addition to data collection, we show that Ken is well suited to anomaly- and event-detection applications. A key challenge in this work is to intelligently exploit spatial correlations across sensor nodes without imposing undue sensor-to-sensor communication burdens to maintain the models. Using traces from two real-world sensor network deployments, we demonstrate that relatively simple models can provide significant communication (and hence energy) savings without undue sacrifice in result quality or frequency. Choosing optimally among even our simple models is NPhard, but our experiments show that a greedy heuristic performs nearly as well as an exhaustive algorithm.

Abstract. In moving object databases, object locations in some multidimensional space depend on time. Previous work focuses mainly on moving object modeling (e.g., using ADTs, temporal logics) and ad hoc query optimization. In this paper we investigate logical properties of moving objects in connection with queries over such objects using tools from differential geometry. In an abstract model, object locations can be described as vectors of continuous functions of time. Using this conceptual model, we examine the logical relationships between moving objects, and between moving objects and (stationary) spatial objects in the database. We characterize these relationships in terms of position, velocity, and acceleration. We show that these fundamental relationships can be used to describe natural queries involving time instants and intervals. Based on this foundation, we develop a concrete data model for moving objects which is an extension of linear constraint databases. We also present a preliminary version of a logical query language for moving object databases. 1

: A large spectrum of applications, such as for instance geographic information systems, planning systems and multimedia applications, take into account time and space. When modeling these spatio-temporal applications, the spatial, temporal but also spatio-temporal aspects (like change or motion) have to be expressed, and a well-understood ontology has to be provided. In this paper, we propose a classification of spatial, temporal and spatio-temporal properties of objects and of relationships between them, in order to guide the designer during the conceptual design phase; we consider both qualitative and quantitative properties. Finally, we discuss the mapping of conceptual descriptions to active databases. Keywords: space - time - motion - change - quantitative - qualitative. 1. INTRODUCTION Spatio-temporal applications deal with situations in which not only atomic (as in the relational model) or complex (as in the object-oriented model), but also spatial, temporal and spatio-tempo...

Abstract. In this paper, a new moving objects database model- Discrete Spatio-Temporal Trajectory Based Moving Objects Database (DSTTMOD) model, is put forward. Trajectories are used to represent dynamic attributes of moving objects, including the past, current, and future location information. Moving objects can submit moving plans of different length according to their moving patterns. Moreover, they can divide the whole moving plan into multiple sections, and submit each section only when it is about to be used. Different moving objects can set up different threshold to trigger location updates. When a location update occurs to a moving object, not only its future trajectory is updated, but also the corresponding index records are adjusted. The model can support three kinds of queries (point queries, range queries, and K-nearest neighbor (KNN) queries) for location information in not only the near future, but also the far future. In order to evaluate the performance of the DSTTMOD model, a prototype system is developed and a series of experiments are conducted which show promising performance. 1

This thesis addresses three important problems related to sensor data processing with the purpose to improve the correctness of results in execution of sensor queries. The first problem focuses on how to schedule updates to maintain the temporal validity of sensor data with minimal workload. The second problem is how to select the right set of sensors for sensor data aggregation to obtain data values that are precise enough to meet the probabilistic requirements of sensor queries. The third problem is how to guarantee the accuracy of the query results without incurring significant update cost in the context of Location Dependent Continuous Query (LDCQ). In the first part, we study the real-time scheduling algorithms for update trans-actions associated with sensor data to minimize CPU utilization. Different from the traditional real-time scheduling approaches which adopt periodic update transac-tion model, we propose a novel algorithm, namely deferrable scheduling algorithm for fixed priority transactions (DS-FP), in which update transactions are scheduled following a sporadic task model. DS-FP exploits the semantics of temporal valid-ity constraint of sensor data by judiciously deferring the sampling times of update transaction jobs as late as possible. The schedulability of the algorithm is examined and a sufficient condition is presented in this thesis. We also present a theoretical analysis of its CPU utilization and prove that DS-FP is optimal for fixed priority schedules in terms of minimizing processor workload. To reduce the on-line scheduling overhead of DS-FP, we further propose two hyperperiod-based approaches with lower scheduling overhead and also satisfying the validity constraint. The first algorithm, namely DEferrable Scheduling with Hyperperiod by Schedule Construction (DESH-SC), searches for a hyperperiod in the DS-FP schedule, and repeats the hyperperiod infinitely. The second algorithm, namely DEferrable Scheduling with Hyperperiod by Schedule Adjustment (DESH-SA), adjusts the DS-FP schedule in an interval so that the adjusted schedule in the interval can be repeated infinitely. In this manner, we reduce the on-line scheduling

Abstract: Tracking moving objects is one of the most common requirements for many location-based applications. The location of a moving object changes continuously but the database location of the moving object cannot update continuously. Modelling of such moving object database should be considered to facilitate study of the performance and design parameters for this database feature. Such study is essential for selecting the optimal solution in order to minimize the implementation of the overhead cost. Location updating strategy for such type of database is the most important criteria. This paper proposed a timed Petri net model based on one of the most common updating strategies, namely the distance updating strategy. In addition, a method for estimating the time needed to update moving object database using the concept of the minimum cycle time in timed Petri nets is presented. This time is the main criterion, which can be used to study the overhead communication cost for moving object database. A typical numerical example is given to demonstrate the advantages of proposed modelling technique.