Zhenwen He

Decomposition tree: a spatio-temporal indexing method for movement big data

Movement is a complex process that evolves through both space and time. Movement data generated by moving objects is a kind of big data, which has been a focus of research in science, technology, economics, and social studies. Movement database is also at the forefront of geographic information science research. Developing efficient access methods for movement data stored in movement databases is of critical importance. Tree-like indexing structures such as the R-tree, Quadtree, Octree are not suitable for indexing multi-dimensional movement data because they all have high space cost of their inner nodes. In addition, it is difficult to use them for parallel access to multi-dimensional movement data because they thereof, are in hierarchical structures, which have severe overlapping problems in high dimensional space. In this paper, we propose a novel access method, the Decomposition Tree (D-tree), for indexing multi-dimensional movement data. The D-tree is a virtual tree without inner nodes, instead, through an encoding method based on integer bit-shifting operation, and can efficiently answer a wide range of queries. Experimental results show that the space cost and query performance of D-tree are superior to its best known competitors.

3D GIS database model for efficient management of large scale underground spatial data

One issue of GIS is management of huge amount of spatial data. Nowadays, acquisition capacity of 3D spatial data with various scales and resolutions is much more convenient due to new 3D acquisition technologies. It is a difficulty that how to efficiently and synthetically organize and manage large scale underground 3D spatial data with contents of continuous distribution and nonuniformity. A 3D GIS data management system framework with three layers is proposed to deal with large scale underground data integrated management. The related key issues and methods include: (1) Multi-scale underground spatial object concept model. (2) 3D spatial database model considering spatial and semantic relationship and corresponding data structure is introduced to support extendable storage environment such as file system, typical commercial database management system and cluster parallel system. (3) True 3D spatial index considering Level Of Detail (LOD). (4) Efficient dynamic dispatch method of great amount of 3D spatial data is adopted based on 3D data content and associated information. (5) 3D Spatial Data Engine (SDE) provides a uniform access interface for file management system, relational database management system and cluster parallel system and other applications. The 3D GIS database model focuses on the key issues of massive underground 3D spatial data provides a new way for the high performance database organization and management of true 3D geological data as well as its integration with the aboveground spatial data.

Design and implementation of log management module in three-dimensional spatial database management system

Log management module plays an important role in the massive spatial database management system. Recently there are few three-dimensional spatial database management systems with pertinent log management sub-system, or the administrator can only use the log, which is attached by database system, to recover the data. To improve this issue, hierarchical structure of the log management module should be taken into account. The overall design of three-dimensional spatial database management system is as follows: three-dimensional spatial database (including database system, file system), three-dimensional spatial data engine, and three-dimensional space management tools. The log management module is designed to solve the problem of complexity and inconvenience of three-dimensional spatial data manipulation records. According to the 3D spatial database management system design, we divided the log management into three layers. The log management system includes monitor layer of three-dimensional spatial databases triggers and file database system operations recording statements, interface layer of log management module of three-dimensional spatial database engine, and graphical user interface layer of log management module of three-dimensional spatial data management tools. Monitor layer design includes database system trigger design. When the database table has been dynamically created, the trigger will be automatically created and record the database operations, then the file system calls function to record the log. Interface layer design includes accessing the database, reading the log table in database and other integrated operations, including query and delete. Graphical user interface layer mainly includes the design of graphical interfaces, the function call form interface layer, and log data operations. In order to work on the layers conveniently, database system trigger and file database systems record are used to record the operation of the database. The log module interface in three-dimensional spatial data engine is used to exchange the data between graphical interface and underlying database. The log module in database management tools can directly provide a graphical interface, but not need to provide access functions to get the log data. This design method has an excellent scalability and flexibility, and can also be made appropriate changes to meet the new needs.

Spatial index technology for multi-scale and large scale spatial data

Large scale three-dimensional spatial data of underground and aboveground entity has features of different shapes, uneven spatial distribution and complicated spatial relationships. At present, the single spatial index structure can not satisfy the large scale three-dimensional spatial data organization and management. Especially for the index of the elements with information retrieval LOD, its efficiency is low. Considering these features, a new spatial index structure named LOD_SKDR tree structure is designed for management of large scale spatial database, which is combined with SKD-Tree, R-Trees and LOD object information. LOD_SKDR tree is a complex multi-level spatial index structure, which utilizes the object number as a bifurcation bound targets to improve the balance of SKD tree. At the same time, the R-Tree index space and texture scheduling of the content through SKD tree are restricted to reduce the time spending of the deletion or insertion and the time of stagnation for scheduling a large number of texture image data. LOD model object information is also used to improve the efficiency of real-time rendering of three-dimensional scenes. Compared with R tree, LOD_SKDR tree has great increase in search efficiency. And considering pre-scheduling of object textures, the effect of its application is more pronounced when spatial data is in large scale.

An effective cache management algorithm of three-dimensional spatial data engine

Traditional substitution algorithms for cache management are more unitary, and basically each algorithm only has correspondingly good effect on certain type of access pattern. Based on analyzing the characteristics of spatial data, a new approach is proposed for cache replacement, considering both frequency and the duration of hit comprehensively. The criterion function of substitution algorithm can be adjusted by statistics of database querying condition combined with the system resource usage, so that the cache would have high performance, and competition for system resources could be reduced. This paper described three sub-algorithms of cache management algorithm, which are substitution algorithm based on hit frequency and duration, cache pool management algorithm and object incidence query algorithm. The cache management algorithm has been implemented on the cache management module of three-dimensional spatial data engine and applied in the three-dimensional data processing of urban planning of Wuhan. It has been proved that the cache management algorithm could be used in multi-level buffering structure and object-oriented spatial database management system to improve the capability of spatial data dispatch.

Three-dimensional spatial indexing method of complicated geological scene

How to create high-performance true three-dimensional spatial index for the complicated geological scene to ensure the effective storage management and dispatching of the TB-class mass complicated geological scene data is one of the key technical issues of the 3DGIS, which needs to be resolved urgently. The inefficient space division and the high rate of index node overlap problems will appear in the continuous nonhomogeneous geological space by using the existing three-dimensional R-tree series spatial indexes. They cannot effectively deal with large-scale complicated geological scene data. In order to resolve these problems, the amount of data evaluating techniques of geology solid and multi-levels indices techniques are the introduction of three-dimensional spatial index. Used three-dimensional space division technology based on the amount of data evaluating technology of geology solid, combined with three-dimensional spatial clustering, spatial scan sorting, fast convex hull algorithm and CSR-Tree packing construction techniques, the multi-level mixed three-dimensional spatial index can be constructed dynamically. It advances the query efficiency by using the topological relationships between the geology solids, realizes the complex geological data management and efficient dispatching, and affords new ideas and new method for the continuous nonhomogeneous three-dimensional space.

A Distributed Indexing Method for Timeline Similarity Query

Timelines have been used for centuries and have become more and more widely used with the development of social media in recent years. Every day, various smart phones and other instruments on the internet of things generate massive data related to time. Most of these data can be managed in the way of timelines. However, it is still a challenge to effectively and efficiently store, query, and process big timeline data, especially the instant recommendation based on timeline similarities. Most existing studies have focused on indexing spatial and interval datasets rather than the timeline dataset. In addition, many of them are designed for a centralized system. A timeline index structure adapting to parallel and distributed computation framework is in urgent need. In this research, we have defined the timeline similarity query and developed a novel timeline index in the distributed system, called the Distributed Triangle Increment Tree (DTI-Tree), to support the similarity query. The DTI-Tree consists of one T-Tree and one or more TI-Trees based on a triangle increment partition strategy with the Apache Spark. Furthermore, we have provided an open source timeline benchmark data generator, named TimelineGenerator, to generate various timeline test datasets for different conditions. The experiments for DTI-Tree’s construction, insertion, deletion, and similarity queries have been executed on a cluster with two benchmark datasets that are generated by TimelineGenerator. The experimental results show that the DTI-tree provides an effective and efficient distributed index solution to big timeline data.

Fractal generator for efficient production of random planar patterns and symbols in digital mapping

Abstract In digital cartography, the automatic generation of random planar patterns and symbols is still an ongoing challenge. Those patterns and symbols of randomness have randomly variated configurations and boundaries, and their generating algorithms are constrained by the shape features, cartographic standards and many other conditions. The fractal geometry offers favorable solutions to simulate random boundaries and patterns. In the work presented in this paper, we used both fractal theory and random Iterated Function Systems (IFS) to develop a method for the automatic generation of random planar patterns and symbols. The marshland and the trough cross-bedding patterns were used as two case studies for the implementation of the method. We first analyzed the morphological characteristics of those two planar patterns. Then we designed algorithms and implementation schemes addressing the features of each pattern. Finally, we ran the algorithms to generate the patterns and symbols, and compared them with the requirements of a few digital cartographic standards. The method presented in this paper has already been deployed in a digital mapping system for practical uses. The flexibility of the method also allows it to be reused and/or adapted in various software platforms for digital mapping.

Multilevel updating method of three-dimensional spatial database

Spatial entities and the relationships between them in real world are changing over time, and the demands from continuous updating of all kinds of data are improving constantly. Therefore, how to maintain the current trend of spatial database becomes increasingly important. However, the spatial data, which exhibit the further feature of massive data, multilevel, concurrent operation of multi-users, make the data updating more difficult which is reflected on inefficiency, insecurity and inconsistency, etc. According to the features of spatial data and problems on updating of spatial database, the multilevel updating method of three-dimensional spatial database is presented based on the systematic analysis of theory and technology of Spatial Data Engine(SDE). In our system, the three-dimensional SDE introduce the SDO_GEOMETRY, which is an abstract data type provided by Oracle Spatial, to store spatial data types. The OCCI (Oracle C++ Calling Interface) is also introduced instead of ODBC or JDBC to access and update the spatial data to improve the efficiency and connection stability of the database system. The multilevel structure of spatial data shows its feature of un-structuring, structuring and object-oriented during the storage of database. Finally, the multilevel spatial data updating method, which is realized in modules of Three-dimensional Spatial Data Management System to schedule and update the massive three-dimensional spatial data, improves the efficiency, interaction and integrity of spatial database.

A transmission architecture for massive 3D geospatial data dynamic dispatch based on ACE

Becoming development trend of GIS, three-dimensional visualization makes the amount of spatial data grown inevitably. Therefore, the study on storage and management of massive three-dimensional geospatial data is necessary and significant. The C/S framework, which consists of massive three-dimensional geospatial database for server and a large number of low-profile PC for clients, has become a basic idea to solve the problem. However, the development of server software witnesses plenty of challenges, such as error detection and recovery, intercurrent control and real time dispatch of spatial data etc. Based on ACE, how to apply the framework and pattern of communications software for building up a powerful application framework and constructing a high-performance server called adaptive communication server of massive three-dimensional geospatial data, is presented. The paper discusses massive three-dimensional data server related technology to achieve a thorough exploration and research, analyses the domain of 3D GIS thoroughly, identifies the problem space and the core design challenges, and proposes a feasible solution. Reducing the difficulty and workload of 3D GIS server development, this solution can have the developer focus on the development of data visualization. In addition, this paper described the object oriented architecture which supports multiple server policy configurations. Being a dynamically ordered structure, it can be ordered systematically these strategies and can be statically and dynamically change their behavior, so that it can provide the most effective strategies for a given software or hardware platform and work load.