The Data Management And Index Technology In Embedded Database For An Intelligent Vehicle Terminal

With the rapid development of intelligent vehicle terminals, the mount of data that needsprocessing increases a lot as the functions of intelligent vehicle terminals become richer andricher. Embedded database, as the most important data storage for application on vehicleterminals, plays the role of ensuring the safety of data, improving data management, andfurther being suitable for the particular properties like lower storage, high real time onembedded systems. In addition, its performance has attracted more and more attention.In this paper, we use vehicle terminals installed on dangerous cargo transport vehicles asthe platform, researched on issues appeared in data management of embedded database, suchas poor real-time data store, large space consumption, and propose an improved databaseindexing. Firstly, The paper studied the characteristics of the vehicle terminal in datamanagement, and the fundamental theory of embedded database, analyzed B-Tree index of itsdrawback space performance used in SQLite database. Secondly, the paper explains theindexing technology like Hash Table, AVL tree, T tree, T-tail tree and hybrid indexingHybrid-TH and proposes, on the basis of above research, a new indexing mechanism--H-T-tail, which not only reduces the space usage, and the query time as well, and meets therequirements of embedded system better. Lastly, the paper finishes the performance testing ofH-T-tail indexing, designs and implements the data management module in intelligent vehicleterminals, improves the architecture of the indexing for SQLite database and compares theperformance of the improved database with the traditional one.The testing results show that, in the case of uniform distribution, The improved indexingH-T-tail is better than B-Tree and Hybrid-TH indexing both in time and space performance.From the experiments on vehicle terminals, it proves that the new indexing has betterperformance in real applications and fits the particular properties for embedded system likelow storage and high real time.