An Architecture for Multiple , Large User-specific Data in a Networked RFID

of Master’s Thesis Academic Year 2007 An Architecture for Multiple, Large User-specific Data in a Networked RFID Summary As RFID technology has been widely adopted, it has been acknowledged that user-specific data, as well as a unique identifier, carried by RFID tags, can be conveniently used. The typical user-specific data stored in an RFID tag is sensor data. There is also an industrial demand to include non-sensor data, such as sales records, repair history, etc., as well as sensor data. It is expected, therefore, that the future information system for networked RFID needs to incorporate heterogeneous userspecific data in an RFID tag in a consistent architecture. There are two fundamental challenges to handle such heterogeneous user-specific data in RFID tag. The first one is the establishment of a logical link between the application and the user-specific data in physical memory in a RFID tag. The other is the efficient retrieval of large user-specific data in RFID tag in practical radio noise environments. An illustrative example of large user-specific data is the retrieval of some kind of logging data in an RFID tag memory. In this thesis, new functions named ”schema resolver” and ”reader controller” at middleware layer are introduced to meet these two challenges in a networked RFID system. The schema resolver provides the link between the object identifier, which identifies the target user-specific data in the application interface, and the schema of RFID tag memory. Since the fundamental drawback of the schema resolver might be the lookup overhead, the overhead is estimated with a simulation based on experimental data. It was revealed that the overhead is equivalent to and in some cases even better than that of the existing method, which exclusively relies on the memory schema information in a tag memory. The retrieval time of a large data from an RFID tag was also examined in an experiment. The result indicated that it took a lot of time for large RFID tag data retrieval. An adaptive block size control mechanism in a reader controller is proposed in this thesis to solve the problem. The mechanism features the automatic reading block length adjustment, based on the retransmission frequency. We confirmed the priority of the proposed method from the mathematical simulation and the experiment.