Research On Main Memory Index For RFID Middleware

RFID(Radio Frequency Identification) is the technology for frequency contactless communication by radio. This is the rapid identification technology which can automatically identify and access data objects. RFID middleware is an important component of RFID systems. It is a bridge between the representative in the physical world of the RFID and various application systems. The current direction of development is to supply variety of RFID middleware platform, and thus RFID middleware technology will be introduced. RFID technology-based system will generate a lot of real-time RFID data, which are transmitted via RFID from the RFID reader to middleware. The particular critical problem for the RFID middleware is how to efficiently store these large amounts of data, to provide an efficient application to the upper access.First, we study common indexing techniques and cache optimization techniques, common indexing techniques, such as Hash, AVL tree, B-tree, B + trees which utilization of storage space is not high, not suitable for the memory index. T Tree improved utilization of storage space, but it did not aware of the cache (Cache) importance under the conditions at the time of the hardware. Cache optimization techniques can be divided into two kinds of active-and passive. Active cache optimization techniques typically require the support of related hardware, and passive caching optimization primarily from the data structure.Secondly, this article analysis in great details the characteristics RFID data and its operation. According to the definition of the EPC Global middleware interface, this article brought a index T + tree structure base on memory-based cache memory technology and optimization technology for RFID middleware memory index of the array. It also provides various operating RFID data algorithm of. This index structure would be more effective to improve memory utilization besides supports all EPC Global standard interface. It in particular, greatly improve the RFID application system commonly used to find the conditions of the event period and the volume range query query efficiency.Finally, Chapter 4 have compared experiments and analysis the three array index of the disk T + tree index structure and the B-tree, AVL tree, the memory T-tree index structure and Zhangduo proposed an index structure suffix_T RFID stream data tree. It also includes improvement of T nodes Cache hit rate experiments. Experiments show that there is certain improvement for the structure of space utilization in the storage, query and update performance.