Research And Improvement Of Anti-collision Algorithms For RFID

With the rapid development of the Internet of Things,Radio Frequency Identification(RFID),one of the four core technologies supporting the Internet of Things,is becoming more and more important.In the RFID system,when multiple tags send their ID data to the reader at the same time,the information between them will interfere with each other,that is,tag collision(or tag collision).This will cause the reader to fail to recognize the tag information properly.Multi-tag conflict not only reduces the identification efficiency of RFID system,but also increases the probability that tags are not read,which leads to a longer identification time of the whole system,which limits the development of RFID technology to a certain extent.Therefore,how to solve the problem of multi-tag collision in RFID system has become one of the important problems.At present,there are two kinds of anti-collision algorithms for RFID tags,one is uncertain anti-collision algorithm based on ALOHA,the other is deterministic anti-collision algorithm based on binary tree.This paper mainly studies deterministic anti-collision algorithm based on binary tree.In this paper,a group-paging for lock-bit dual prefix probe scheme(GP-LDPPS)is proposed.The algorithm uses Manchester encoding to find the collision bits,and then the reader sends the lock instructions to the tag.The tag receives the lock instructions to extract the collision bits of its ID to form a new ID,and then groups them according to the information of the collision bits of the new ID.The reader records the grouping situation through an array of Ac.After grouping,each group of tags is identified by means of double prefix probes.By effectively grouping tags,the number of collision slots is reduced,and the number of idle slots and the amount of data transmission are reduced.Then,according to the performance indicators of GP-LDPPS,such as total slot number,transmission delay,recognition efficiency and throughput,theoretical derivation is made,and three aspects of GP-LDPPS are simulated and analyzed.The simulation results show that the average throughput of GP-LDPPS algorithm reaches 96.5%,and the recognition efficiency is about 0.51.It effectively reduces the data transmission,improves the recognition efficiency,reduces the number of queries,and enables the tags to be identified more quickly and successfully.On the other hand,based on the multi-cycle Collision Tree(MCT)algorithm,this paper proposes a back-locking backoff Collision Tree anti-collision algorithm(BLBOCT).According to Manchester coding idea,this algorithm determines the location of tag collision,extracts collision bit information by sending lock instructions,and only sends collision bit in the future data transmission process.This not only reduces the collision time slot,but also reduces the total amount of data transmitted.Then,this paper analyses the performance of BLBOCT algorithm,and simulates the algorithm in three aspects: total slots,throughput and recognition efficiency.The results show that the throughput of BLBOCT algorithm is about 96.4% and the recognition efficiency is about 0.55.BLBOCT algorithm further reduces the data transmission,improves the system throughput,and also improves the overall recognition efficiency of the system.