The RFID system consists of a large amount of RFID tags and a reader. It is expected to create new services that cannot be achieved by the traditional bar code system. Specifically, in a distribution system, a high-speed and continuous readout method of ID information and data from a large amount of RFID tags embedded in products is required to reduce costs and time. A reader communicates with RFID tags via a wireless communication. So, there is a problem that a reader cannot simultaneously read information from multiple RFID tags at once because a collision occurs among responses from them. In this thesis, we propose novel methods, called Response Probability Control Method (RPCM), to read information from a large amount of RFID tags quickly and continuously by avoiding collisions. In RPCM, a reader firstly sends an ID request to RFID tags in its access area. To reduce collisions, the reader appropriately controls the probability that a tag responds to the ID request. The response probability is determined by the estimated number of tags in the access area. Mathematical analysis and simulation results show that RPCM can accomplish the readout efficiency 1.6 times higher than the traditional method on the average where there are a thousand of RFID tags whose IDs are 128 bits. In addition, by considering that the number of bits required to distinguish tags in the access area is often smaller than that of the original ID, we use part of the original ID, called temporal ID to distinguish RFID tags. The reader first gathers the temporal ID according to RPCM. Then, it collects the latter part of ID, called remaining ID, or data by designating the corresponding tag with the temporal ID. In this way, we can reduce the amount of lost data and improve readout efficiency. Since the degree of the performance improvement is affected by the size of the original ID and the number of tags, we investigate the feasible area of the system using the temporal ID. Mathematical analysis shows that the degree of the performance improvement becomes large if the size of an original ID is large and the size of a temporal ID and that of data are small.
Wataru Nagate, High-Speed Information Readout Method from a Large Amount of RFID Tags, Master's Thesis, March 2007.
@mastersthesis{nagate07mthesis,
author = "Nagate, Wataru",
title = "High-{{Speed Information Readout Method}} from a {{Large Amount}} of {{RFID Tags}}",
year = "2007",
month = "March",
school = "Osaka University"
}