Study On Tag Anti-collision Algorithms In RFID Systems

In the Internet of Things system,information collection is the first step in data processing and control.Commonly methods used for data acquisition include barcodes,sensors,video surveillance,and radio frequency identification(RFID).Among them,RFID is one of the most important technologies adopted for automatic identification due to its non-contact,tag content modifiability,batch read and other advantages.At present,RFID technology is widely used in fields such as intelligent transportation.intelligent logistics,and intelligent manufacturing which need to realize large-scale automatic identification of objects.The RFID system includes readers,tags,and applications.Under normal circumstances,the reader actively challenges the tag information,and the tag passively responds to the challenge and returns the information.With the widespread application of RFID systems in the Internet of Things,not only tags with low price and fast response are needed,but also readers with fast reading speed,high recognition efficiency,high accuracy,and low tag loss rate.Improving reader recognition efficiency,speed,and accuracy based on existing storage capacity,computing power,and computing speed of RFID reader is a key issue to be urgently addressed in the development of RFID technology.It is of great significance to further expand application of RFID in the field of intelligent IoT and intelligent information collection.This dissertation focuses on the shortcomings of the DFSA algorithm in the application process,and proposes RFID tag anti-collision algorithms for the problem of tag number estimation and dynamic tag arrival rate.It conducts in-depth research on capture effects and online prediction of tag arrival rate.The main contents include the following aspects.1.Based on the maximum a posterior probability and capture effect,an RFID anti-collision algorithm is constructed to improve the prediction accuracy of the number of tags and the recognition speed of the RFID system.Aiming at the situation that the number of tags in each frame is unchanged in a static RFID system,a DFSA-based anti-collision algorithm MAPCE is proposed based on the capture effect phenomenon and the maximum posterior probability method.First,the maximum a posterior probability method that takes the capture effect into account is used to estimate the number of tags.Then,the frame length is set according to the frame length setting method of the unequal-length slot.Finally,a tag recognition operation is performed.This algorithm takes the capture effect of the signal into account when it propagates in space,and can identify a tag according to the energy relationship in the event of a tag collision,which greatly improves the efficiency and speed of tag recognition.2.Based on the recognition process of the mobile RFID system,a model of tag movement in the recognition area is established,and a backward correction method is proposed to predict the arrival rate of the tags,and the conflict between the new and old tags in the recognition area is discussed.In the mobile RFID system,the tags enter the reader’s identification range continuously.and at the same time,the tags move out of the identification area.Based on the dynamic arrival model of tag movement,the backward revision tag arrival rate method is adopted,and the priority of the new and old tags is established.A mobile tag anti-collision algorithm RPA based on round priority is proposed.The RPA algorithm uses two priorities to completly eliminate the conflict between new and old tags,and reduce the probability of collision and the rate of missed tags.3.Based on the gray model GM(1,1),an online prediction model for label arrival rate is constructed,and the initial value of the differential equation is modified using a weighted method.The sliding window mechanism is used to implement sequence update and isometric modeling,and then to achieve continuous online prediction of label arrival rate.In mobile RFID systems,the tag arrival rate is unknown.How to accurately predict the tag arrival rate is the key to accurately set the reader frame length.The GM(1,1)model can use at least 4 data to complete dynamic online prediction,which meets the characteristics of small storage capacity and small computing capacity of RFID readers.The weighted method is used to improve the initial value of the gray differential equation,and the sliding window mechanism is used to ensure that the modeling length is 4.Therefore,the weighted GM(1,1)algorithm WGMSW(1,1)based on the sliding window is proposed,which greatly improves label arrival Rate prediction accuracy.4.Using WGMSW(1,1)to predict the tag arrival rate and combining with the DFSA basic algorithm,a mobile RFID system anti-collision protocol GM-DFSA is proposed.Simulation analysis of GM-DFSA’s recognition speed,recognition efficiency and tag loss rate show that this protocol has better performance than the Q algorithm of EPC C1G2 and it can be applied to mobile RFID system to get rapidly identification.