What's SuperMap SDX+?

SuperMap Spatial Database eXtension (SuperMap SDX+) is a geospatial database eXtension, enabling all SuperMap GIS products with the ability of accessing geospatial database, as well as the function of storing, indexing, reading and updating spatial data with DBMS, Most of SuperMap products, for example, SuperMap Deskpro, SuperMap Objects and SuperMap IS .NET contain built-in geospatial database eXtension SuperMap SDX+, so that they can manage spatial data with DBMS internally.

Data storage and its accessibility are critical for the overall performance of an entire GIS application; therefore, the functionality and performance of the database engine have a fundamental effect on the integrity and efficiency of the GIS application. Through years’ research and improvement, SuperMap SDX+ has become a reliable and high performance geospatial database eXtension, by which, spatial geometric object data and its attribute data can be saved integratively into various relational databases, and varieties of maintenance work, for example, index maintenance, appending, updating and deleting operations can be carried out. Query can be performed based on either attribute or spatial condition to return required data. In addition, many advanced functions such as long transaction, version and topological relationship maintenance are also provided.

SuperMap SDX+ supports most of the popular commercial database platforms, such as Oracle, SQL Server, Sybase, DB2 etc., which can be run on multiple operating systems. Therefore, multi-node cluster can be constructed among homogenous database while distributed cluster can be constructed on heterogeneous databases and operating systems. Moreover, SuperMap SDX+ is developed by using standard C++ programming language, making cross platform development possible, so that flexible spatial data accessibility and management can be achieved on more operating systems (including Linux, UNIX etc.).

Figure 1 SuperMap SDX+ spatial database eXtension

Technical Features

There is a trend in GIS to use large relational databases for integral management of spatial data and its attribute data. Spatial database technique has its advantages in many aspects as follows: Massive data management capability, integral storage of spatial data and attribute data, multi-user concurrent access (read & write), complete access permission control and data security mechanism etc.

More and more large- and middle- scale GIS projects take spatial database technology as the solution for their spatial data storage and management. As the spatial database eXtension of SuperMap GIS, SuperMap SDX+ is a very important part of SuperMap GIS software. It adopts advanced spatial database storage technology, indexing technology and query technology, and has the integrated spatial database management ability of “integrating spatial and attribute data”, “integrating vector and raster data” and “integrating spatial information and business information”, making it qualified for large-scale GIS projects.

Three distinct features of SuperMap SDX+ are identified by a series of applications and tests as follows:

　　◆ Easy installation and deployment and make full use of database technology.
　　◆ High performance massive data access and management.
　　◆ Perfect data model sufficing the requirements of various large GIS applications.

Easy Installation and Deployment

　　◆ Easy configuration

SuperMap SDX+ requires no configuration of complex spatial database servers, all of its deployment and management adopt standard database operations, which means that no more extra database technique is needed so that an average database administrator can handle this task.

Standard database management procedure utilizes DBMS’s strengths in data security control, permission control etc. without security vulnerability caused by some special account types. Access to database via standard database protocol can satisfy smooth data exchange over the web. What’s more, SuperMap SDX+ manipulates database via standard SQL syntax so that the performance of the server can be fully leveraged while users on client side can perform query and maintenance tasks based on standard database procedures.

　　◆ Full support for popular commercial relational database platforms

SuperMap SDX+ supports mainstream commercial relational database platforms including: Oracle, Oracle Spatial, SQL Server, DB2 and Sybase as well as native DM and Kingbase.

As SuperMap SDX+ has encapsulated the interfaces for accessing these database platforms making them transparent for users, they do not need either to care what kind of server hardware or operating system they are running on or to know background database type, server type and operating system etc., access and management of the data can be proceeded via similar interfaces, reducing the difficulty of database configuration, deployment and application.

　　◆ Support for Oracle RAC

RAC (Real Application Cluster) is one of the key functionalities of Oracle database 10g Enterprise Edition with shared cache architecture. RAC technology perfectly solves two fundamental problems: Firstly, other servers can automatically take over the tasks of the server which breaks down; more importantly, all servers within the cluster are concurrently running to greatly enhance the capability of handling requests from client sides.

In January, 2005, SuperMap together with Oracle and HP performed the compatibility tests for SuperMap GIS with Oracle 10g and Unix/Linux systems. The result shows that SuperMap SDX+ has a perfect compatibility with Oracle 10g on Unix/Linux platform and it fully supports the most advanced RAC technology of Oracle 10g, indicating SuperMap SDX+ has a strong edge in compatibility and massive data handling. Based on SuperMap SDX+ and Oracle RAC technology, SuperMap GIS can be applied in large-scale GIS applications with reliable and high performance services.

Figure 2 Oracle RAC technology

High Performance Access and Management of Massive Spatial Data

　　◆ Hybrid multilevel indexing technique

Indexing technique is a key technique in spatial database engine, which directly affects the efficiency of data access and query. However, any indexing technique has its disadvantage, and no single indexing technique can satisfy the requirements of data access to hybrid storage model of massive image data and vector data. Therefore, SuperMap comes up with hybrid multilevel indexing technique, which fuses multi-layer grid, quad tree and R tree indexing methodologies, so that advantages of the three can be harnessed.

As to the spatial data with standard sheets, SuperMap also supports creating library index according to specified field or spatial extent which coupled with file cache technique can provide great access ability.

Moreover, it is also possible to create attribute field for vector dataset as well as delete field index. Users are allowed to maintain the index of attribute data on the SuperMap environment to speed up query and access to the attribute data.

　　◆ File cache technique

File cache provides an intelligent distributed storage scheme for balancing the load of a server and web, and improving overall performance. With file cache enabled, the local cache is firstly searched to see if there is the latest version of the required data when an application intends to access the data in the spatial database. If there is no related cached file or the cached data is not the latest one, the required data will be read from the server and be used to update the local cache, so that next time the data can be directly accessed in local cache; if the latest version of the required data is available in local cache, then there is no need to request data from the server while browsing and analysis can be performed via direct access to the local cache. Through this scheme, server and network load can be tremendously reduced while overall performance is greatly improved.

In order to save disk space and increase flexibility, there is no space index configured in the file cache, so it is not appropriate to enable file cache for a dataset with many records. But local cache can be jointly used with library index, e.g. the data is cached according to the sheet it belongs to if the library index has already been built for the data. In this way, cache size will not be very large and fast data loading and update will be ensured. So a combination of library index and file cache will be an optimal solution for the vector dataset with many records.

And of course, file cache mechanism will occupy the disk space of client-side, SuperMap SDX+ employs highly efficient compression algorithms to reduce the space occupied by the cache file; while on the other side, the cache quota function is offered to control the total possible size occupied by the cache.

　　◆  Lossless and lossy data compression

With the development of new data acquisition technology, temporal and spatial resolution of GIS data increase constantly, so does its data volume, causing a severe stress on network bandwidth, transmission rate and storage space. Compressing data is an efficient way to save storage space, bandwidth, and increase data transmission rate; besides, compressed data can be used in secrecy communication to improve data security and system reliability.

SuperMap SDX+ provides both lossless and lossy compression techniques for vector and raster data. Lossless compression is a compression of statistical redundancy, when decompressed, there is no data loss and the original data will be restored. As compression rate is limited by statistical redundancy, so compression rate of spatial data is usually from 2:1 to 5:1. This type of compression applies to industries or applications with demand for high precision data. Due to the limitation of compression rate, lossless compression can not completely solve the problem of massive spatial data storage and transmission, thus, it has a narrow usability.

To obtain higher compression rate, SuperMap SDX+ provides lossy compression for vector and raster data. Lossy compression technique can implement much more compression rate than lossless compression, but there is a little information lost. Although not all original data can be restored when decompressed, lost data has little effect on precision, therefore, it has a wide application in most GIS projects. However, it should be noted that lossy compression technique for vector data differs from that for raster one. Lossy compression for vector data takes the advantage of the adjacent relationships between vertices of spatial objects to perform frequency division encoding of the coordinates of the vertices; while lossy compression for raster data is similar to image compression in multimedia applications, the main compression algorithm including Run-Length Encoding, Discrete Cosine Transform and wavelet transform.
The build-in compression technique can intelligently adjust compression parameters according to the features of spatial data, reducing compression procedures while high compression rate and efficiency can still be promised.