Research On Radio Frequency Identification_RFID Tag Anti-collision Algorithm

RFID(Radio Frequency Identification) is an automatic identification and data acquisition techniques, through radio waves and radio frequency identification technology enables non-contact identification, and communicate with multiple targets simultaneously in real time. Due to its strong penetrability, high reliability and multi-target identification, it is widely used in medical, logistics and other fields.Anti-collision technology is one of the key technologies of RFID. When RFID system works, if more than one label or reader in the same work area, they will interfere with each other when identifying, the technique to solve this interference is called anti-collision technology. Anti-collision problem is divided into reader anti-collision and tag anti-collision problem, and tag anti-collision problem is one of the research hotspots and difficulties.The subject comes from the National Natural Science Foundation Project "IOT Pan localization theories and key technologies", the project number is61303207. This paper introduces the working principle of radio frequency identification(RFID) technology, composition and classification of system, and conducts an analysis of the key technology of RFID, finally, presents the key research technology—anti-collision technology. We detail the two typical types of anti-collision algorithm, including deterministic algorithm based on binary tree, and probabilistic algorithms based on ALOHA, we also analysis the advantages and disadvantages of the two types of algorithm.This paper proposes two improved algorithms based on ALOHA. One of is an estimated dynamic frame slotted ALOHA algorithm, the algorithm uses "projected+parity" approach in terms of the label estimates, firstly, give the label an estimated value, then use the next frame identified tag to check and fix it, consequently achieve dynamic estimation of label. Simulation results show that estimated dynamic frame slotted algorithm has high throughput, and shorten the recognition time effectively. The other is a modified Q value anti-collision algorithm, the algorithm is based on EPCglobal Class-1Generation-2(EPC C1G2) Standard, and the Q value adjustment mechanism has been improved, the reader does not have to start to adjust Q value until after the end of a frame, thus slot utilization can be improved. EPC standard does not specify the valve of adjustment step C, the improved algorithm introduces the concept of threshold, to perform theoretical analysis and simulation. On the basis of solving threshold value, we deduce the C value varying with Q value. By comparison, the throughput of improved algorithm can be maintained at0.4, the total number of slots required for the whole recognition process increase non-linearly with the growth of labels, which has a significant advantage especially when the number of labels are large.