Radio Frequency IDentification (RFID) is a generic term for technologies which use RF waves to identify, track, or categorize any object. One of the research areas in RFID systems is the tag anticollision protocol; how to reduce identification time with a given number of tags in the field of an RFID reader. There are two types of tag anti-collision protocols for RFID systems: tree based algorithms and slotted ALOHA based algorithms. Since the tree based tag anticollision protocols achieve 100% read rate, we consider how to improve the performances of the tree based RFID tag anti-collision protocols.
This thesis proposes three methods for fast tag identification: bislotted tree based RFID tag anti-collision protocols, query tree based reservation, and the combining method of them. First of all, bi-slotted tree based RFID tag anti-collision protocols, bi-slotted query tree algorithm (BSQTA) and bi-slotted collision tracking tree algorithm (BSCTTA), decrease in both prefix overhead and iteration overhead by the time-divided responses depending on whether the collided bit is ``0`` or ``1``. Next, query tree based reservation (RN16QTA: the 16-bit random number aided query tree algorithm) diminishes in the tree depth on the identification process for decreasing the identification time. The query tree based reservation applies the characteristics of temporary IDs in EPC Class 1 Gen. 2 protocol to the query tree algorithm, which is for assigning slots to transmit tag IDs, instead of the real tag IDs. Finally, bi-slotted query tree algorithm takes the advantages of both BSQTA and RN16QTA. sequently, these proposed methods can be used for enhancing the identification speed of RFID systems.