The Uhf Rfid Reader Anti-collision Algorithm

Radio Frequency Identification(RFID) is a non-contact automatic identification technology. Because of their long identification distance, strong penetrating ability, multi-object recognition, the technology is now widely used in automation, transportation,control and management, product services and certificates Anti-forgey areas. It is one of current research hot technologies of"Internet of Things". When more than one tag send data at the same time in the scope of the reader, collisions will happen. The results will lead to transmission failure. Therefore, anti-collision algorithm is needed to avoid or reduce the collision, thereby enhancing tag identification performance of RFID System.At present, the research of anti-collision algorithm in RFID is focused on the Aloha slot anti-collision algorithm and binary tree search anti-collision algorithm. However, aloha slots anti-collision algorithm exist the phenomenon that a label can never be identified. Binary search algorithm can avoid this problem, but when the relatively large number of tags are needed to identify, the reader needs a long identification cycle. Thus the RFID reader efficiency is affected.This paper can be divided into two parts. The first part is mainly applied the Intel R1000 UHF reader to conduct network test. The reader's inherent algorithm is based on ISO/IEC 18000-6C standard algorithms. But it not fully realizes the standard algorithm. The paper analyze the defects in the inherent algorithm. By the simulation of standard algorithm based on ns2, the improvement of inherent algorithms is proposed. Based on the ISO/IEC 18000-6C UHF RFID standard algorithms, This paper proposed a new Slot-based Partial Competitive(SPC) anti-collision Algorithm, which introduces dynamic binary tree search technology into the competitive collision avoidance mechanism for the first time. Moreover, the SPC algorithm applies the specific technologies to idle slots and collision slots. The performances of the dynamic binary tree search, 18000-6C and the SPC algorithm are compared via NS-2 Multi-tag identification simulation experiments. Simulation results indicate that the SPC algorithm is able to not only reduce the recognition time by 30% and 20% compared with the binary tree search algorithm and the 18000-6C algorithm, respectively, but also improve the recognition rate by 35% compared with 18000-6C algorithm. The algorithm has a good value.