Research On Multiple Tags Anti-collision Algorithm Of The Radio Frequency Identification Technique

Radio Frequency Identification (RFID) is a non-contact, automatic data collection technology, which employs electromagnetic waves as the transmission medium to achieve two-way transmission and automatic identification. With many advantages, such as the strong abilities of identifying multi-object at the same time, non-contact reading and writing and environmental adaptability, RFID, which can be used to track and identify almost all entity objects, has been widely deployed in many fields.The ability of identifying multi-object at the same time makes RFID technology stand out from the other automatic identification technologies. The anti-collision algorithm is the key technology and a prerequisite for the realization of multi-object identification. Currently, researches on the anti-collision algorithm mainly focus on the nondeterministic algorithms based on Aloha and the deterministic algorithms based on binary search. In the nondeterministic algorithms based on Aloha, the fatal weakness that the tags in the active range of Reader will probably never be identified will result in the"starving"phenomenon. Although the binary search algorithm can solve this problem, it also has the shortcoming of the long circle for identification.Based on the research of the current binary anti-collision algorithm, this thesis proposes an improved binary anti-collision algorithm and a grouping binary anti-collision algorithm. In the improved binary anti-collision algorithm, the Reader firstly provides the information about the collision bits detected for all the tags in the active range. Then, the tags lock their own corresponding bits, and make the locked binary collision bits as the new identification objects in collision detecting. The handling process of the Reader employs the back strategy, which means that after a tag is recognized, the Reader doesn't search from the beginning, but turns back to the last collision node.The grouping binary anti-collision algorithm mainly aims at the situation of dense distribution of tags. Firstly, the Reader sends a request command within its active range, and with all tags responding to the command, the Reader will send the command of locking the binary collision bits according to the result of decoding. Then, according to the collision bits the Reader estimates the number of tags, which determines the number of groups. Finally, by adjusting the transmit power and changing the transmit antenna gain, the Reader changes the range of reading and writing in the order from the near to the distant, and uses the improved algorithm to deal with the tags in each group until to the margin of reading range. As the tags are locked twice in this algorithm, the redundant information transmitted can be reduced.The mathematical analysis and simulation show that the improved algorithm and grouping algorithm proposed in this thesis have less redundant information, higher system throughout and efficiency of identification. In the situation of dense distribution of tags, the grouping algorithm is more efficient than the improved algorithm in system throughout.