Research On Indexing Technology Of XML Wireless Data Broadcast

With the rapid development of technology, mobile devices are widely used for their portability and high performance. High speed wireless network (i.e.3G, WIFI and etc) also develops rapidly with its cover area wider and wider. The users used to retrieve the data from PC, but nowadays they rather like to use the mobile devices, such as Smart Phones and Tablet PC. As an efficient way to public information, wirless data broadcast enjoys a popular usage in wireless area for its brandwidth-efficiency, energy-efficiency and scalability.XML (short for eXtensive Markup Language) is widely used in every field since introduced by W3C Organization. For the better performance of data exchange, the applications in wireless environment also use XML as the data transfer format. However, because of the differences between wireless and traditional cable network, the existing approaches can’t be simply used in wireless area. Therefore, XML data broadcast technology in wireless environment graduately becomes the research hotspot.In this thesis, we mainly focus on the research of indexing technology in wireless data broadcast environment. By downloading the index, users can know when their interested documents arrive. The devices or applications can tune into doze mode to save energy and back to active mode to download the data when the documents arrive. XML document is different from the data item in tradiational wireless broadcast. XML document is semi-structured and the query from the users is always like XPath with path structure.As for the semi-structured XML, in this thesis, we focus on the index construction with users’ queries in on-demand wireless broadcast. Building the index with users’ queries can be the best reflection the demand of them, they just need to download the minimal size of index node to find the match documents information. At the same time, we introduce a bit-based compressing strategy to further reduce the index size. We further propose a new index extended from the aboved method to support twig queries.We carry varies of experiments to show that generating index by users’queries can effectively improve the performance of the on-demand wireless broadcast.