Zhexuan Song

K-Nearest Neighbor Search for Moving Query Point

This paper addresses the problem of finding k nearest neighbors for moving query point (we call it k-NNMP). It is an important issue in both mobile computing research and real-life applications. The problem assumes that the query point is not static, as in k-nearest neighbor problem, but varies its position over time. In this paper, four different methods are proposed for solving the problem. Discussion about the parameters affecting the performance of the algorithms is also presented. A sequence of experiments with both synthetic and real point data sets are studied. In the experiments, our algorithms always outperform the existing ones by fetching 70% less disk pages. In some settings, the saving can be as much as one order of magnitude.

Hashing Moving Objects

In many real-life applications, objects need to be both spatially and temporally referenced. With the advancements of wireless communication and positioning technologies, the demand for storing and indexing moving objects, which are the objects continuously changing their locations, in database systems rises. However, current static spatial index structures are not well suited for handling large volume of moving objects due to massive and complex database update operations. In this paper, we propose a new idea based on hashing technique: using buckets to hold moving objects. The database does not make any change until an object moves into a new bucket; therefore, the database update cost is greatly reduced. Then, we extend the design of existing system structure by inserting a filter layer between the position information collectors and the database. Based on the new system structure, we also present two indexing methods. Finally, different aspects of our indexing techniques are evaluated.

SEB-tree: An Approach to Index Continuously Moving Objects

Recently, the requirement for storing the locations of continuously moving objects arises in many applications. The paper extends our previous work on zoning based index updating policy [1]. In the paper, we give the data format of object movement under the policy. Then, we propose the SEB-tree (Start/End time stamp B-tree). This index structure has fast insertion and query algorithm, and it outperforms the existing structures in the experimental evaluations.

Using Hilbert curve in image storing and retrieving

In this paper, we propose a method to accelerate the speed of subset query on uncompressed images. First, we change the method to store images: the pixels of images are stored on the disk in the Hilbert order instead of row-wise order that is used in traditional methods. After studying the properties of the Hilbert curve, we give a new algorithm which greatly reduces the number of data segments in subset query range. Although, we have to retrieve more data than necessary, because the speed of sequential readings is much faster than the speed of random access readings, it takes about 10% less elapsed time in our algorithm than in the traditional algorithms to execute the subset queries. In some systems, the saving is as much as 90%.

MOCHA: a database middleware system featuring automatic deployment of application-specific functionality

Introduction MOCHA1 is a novel database middleware system designed to interconnect data sources distributed over a wide area network. MOCHA is built around the notion that the middleware for a large-scale distributed environment should be selfextensible. This means that new application-specific data types and query operators needed for query processing are deployed to remote sites in automatic fashion by the middleware system itself. In MOCHA, this is realized by shipping Java classes implementing these types or operators to the remote sites, where they can be used to manipulate the data of interest. All these Java classes are first stored in one or more code repositories from which MOCHA later retrieves and deploys them on a “need-to-do” basis. A major goal behind this idea of automatic code deployment is to fulfill the need for application-specific processing components at remote sites that do not provide them. MOCHA capitalizes on its ability to automatically deploy code to provide an efficient query processing service. By shipping code for query operators, MOCHA can produce efficient plans that place the execution of powerful data-reducing operators (filters) on the data sources. Examples of such operators are aggregates, predicates and data mining operators, which return a much smaller abstraction of the original data. In contrast, datainflating operators that produce results larger that their arguments are evaluated near the client. Since in many cases, the code being shipped is smaller than the data sets, automatic code deployment facilitates query optimization based on data movement reduction, which can greatly reduce query execution time. The architecture for MOCHA consists of four major components: Client Application an applet, servlet or Java ap-