Research On RFID Muti-tag Anti-collision Algorithm Based On Adaptive Pruning Query Tree

The number of terminal devices rapidly grows,and diverse data is emerging.How to realize the ubiquitous connection between things and things or things and people is a pilot issue for every country to step into the era of universal perception nowadays.As one of the core technologies in the sensing layer of IoT(Internet of Things),RFID(Radio Frequency Identification)is a wireless communication technology.RFID tags which attach to the inside or the surface of objects effectively "communicate" with readers through radio frequency signal.The Internet is used to control,store,track,and manage the collection and processing of related data.These can help realize the effective information interaction and identification.In the future,there must be a large number of tags in the IoT.How to identify each tag with high precision and high speed in the shortest time is a major challenge for the rapid development of IoT technology.In the process of RFID tag identification,tags and readers share the same channel.If multiple tags respond at the same time,it can cause collision and lead to serious identification delay.Therefore,developing effective anti-collision strategies to improve the performance of RFID systems has an important significance.In this regard,in order to improve the system identification efficiency and completely avoid the occurrence of idle timeslots,this thesis innovatively proposes two new RFID multi-tag anti-collision algorithms based on adaptive pruning query tree.According to the characteristics of collision prefixes obtained by the reader through decoding in each tag response process and adopting two different double grouping mechanisms,the two proposed algorithms can accurately group tags and generate valid query prefixes.Firstly,an adaptive bit arbitration query tree anti-collision algorithm based on 4-ary tree(FBAQT)is proposed.The XOR grouping mechanism and automatic identification mechanism are used to generate accurate query prefixes,which can accurately group tags and get rid of all idle timeslots.The performance analysis and simulation experiments are carried out in four aspects: total number of query timeslots,throughput,communication complexity and transmission delay.The results show that this algorithm has better performance than the original related algorithms,significantly reducing the number of queries and making the throughput more stable.The throughput of the algorithm maintains at around 0.605.Secondly,a muilt-tag anti-collision algorithm based on 8-ary adaptive pruning query tree(EAPT)is proposed.The concept of collision coefficient is proposed for the first time.This algorithm only focuses on the three consecutive bits of the decoded information obtained by the reader in each query timeslot which start from the highest collision bit and exists two query mechanisms.According to the ratio of the number of collision bits in the bit string to the total number of the bits in the bit string to determine which mechanism is used to accurately group tags in the next step.Similarly,through theoretical analysis and simulation experiments on four performance indicators,it is proved that this algorithm can significantly reduce the total number of timeslots,completely avoiding the generation of idle timeslots.The throughput is maintained at around 0.625,further improving the system performance.At the same time,in the simulation experiments,by changing the number of tags and the length of tag IDs,it is verified that EAPT algorithm is not restricted by these two factors and still has good identification performance.Through comparison of performance analysis and simulation experiments,this thesis proves that the two proposed anti-collision strategies can guarantee excellent system identification performance in the future IoT application environment under the case of existing massive tags.