A Study Of RFID Data Management Technology Based On Analysis Of Movement Graph

These years, RFID (Radio Frequency Identification) has developed rapidly. As the cost of reader and electronic tags declines and the size of electronic tags get smaller, the accuracy of identifying the tags get better. Some retailers (such as Wal. Mart and Target) have begun their deployment of RFID system in their warehouses and logistics center, and require suppliers paste small RFID tags on the package. When tracking the items with tags, RFID system will produce mass explosive data. So how to manage and store the mass dynamic data effectively becomes important, and it will affect RFID's further development and application.According to analyzing the process for producing data, the difference from traditional data, and the disadvantages of managing data, this paper puts forward a data analysis model based on analyzing item's movement. We divide the movement map into several sub maps by the key points which are on the paths of the majority of items. These sub division are like trees whose root are those key points. The structure of tree is easy to code and query, so we generate sub maps'path tree based on those key points and gain a path code by prime number encoding method. We can make use of the Chinese Remainder Theorem to find the sequence coding the path. Because path code can identify a path uniquely, the sequence coding can locate their position orderly.Thus, it gives the storage model which is based on the path-based coding and sequence coding upon the analysis model. The code of the movement path for the item and the sequence code are stored in the table PathTablle, the coding information for position prime is stored in table LocationTable, and the movement of information items are merged into an item and stored in table the MoveTable, which compress the data greatly for table MapTable mapping the conclusion relationship of items which move together. As table MoveTable recodes the movement information of goods, what we just need to do is relating table PathTable to gain the movement path. Moreover, the separation between table MoveTable and PathTable make the query faster. So this store model can support tracking and querying better. At last, through the analysis for RFID, it gives a query model based on aggregation. The separation between storage model and query model form layered RFID data management system. Finally, because data is stored to a relational database, we give SQL statements which change from the query model.