Study And Realization On Key Technologies Of An Embedded Database

After the IT network technology, the embedded system is another development direction of rising technology. Due to its small volume, strong performance, low absorption, high reliability and special application, the embedded system has been widely used in the fields of national defence, electronic consumption, network communication, industry control and equipment manufacturing.The study on key technologies of embedded database is the research heat in the embedded system field. As research and development of the project "An embedded automatic controller for pockets-opening machine", this paper deeply study part of the key technologies and methods of embedded database system, mainly including the processing ability of a great deal of data, how to transplant Berkeley DB to embedded Linux, and XML parser technologies of an embedded database. In a word, it is a good example for the research and design of an embedded database and its XML parser for transfering and sharing international information. Meanwhile, it is also a practical task for the design and development of embedded system.The main contributions of this paper are summarized as follows:(1) Due to B~+-trees' great shortcoming of low storage utilization and its weak processing ability of mass data in embedded database Berkeley DB, this paper presents an improved B~+-trees method to guarantees high storage utilization on the base of analyzing Hash, Queue, B~+-trees' data structure, performance and their algorithmic flow. After that, the paper gives the comparison results of the improved B~+-trees algorithm to the conventional B~+-trees algorithm through running test suite in Berkeley DB and the experimental results show that the improved B~+-trees method can attain higher storage utilization at the expense of more execution time than B~+-trees, when the variable N is larger, the effect is more distinct.(2) Simple embedded system could use file system as its information management tool, but for sophisticated embedded system, an excellent embedded database will be the best choice. This paper complete an embedded database system software "file manager" which provides basic function of a database system by the API interface of Berkeley DB which has implemented the improved B~+-trees method on embedded Linux.(3) XML file need an XML parser to parse and abstract its content and structure. As Berkeley DB is lack of an interface which can process diversification data with complex structure, the paper presents a method which is to use both DOM and SAX algorithms to implement the XML parser of Berkeley DB to meet the high efficiency and low cost needs of date exchanging interface of Berkeley DB in platform-independent, operating system-independent environment on the base of analyzing DOM, SAX, JDOM, D0M4J algorithms.(4) According to the common parse flow of XML file, this paper design and implement an XML parser for the embedded database Berkeley DB to provide the function of XML file reader which makes some contributions to provide an interface for data exchanging and sharing facing Internet.(5) Our country is a big garment manufacturing and exporting country, but not strong in it Compared with developed countries, out of date of the garment manufacturing machine is one of the critical reasons. This paper applies embedded and computer technologies to design and develop the project "An embedded automatic controller for pockets-opening machine" on the base of developed manufacturing concept. The controller should support online programming, store and manage more than five hundred suit of procedure and remote control. Now the project has been completed on the whole, and also is validated in the garment manufacturing site.