The Research Of Tag Anti-collision Based On The Dynamic Frame Slotted

With the pace of the informatization advancement and application, RFID recognitionsystem has been more and more applied to related human life closely in all aspects. Otherwise, itis still the key factor of restricting label recognition rate to solve the collision problem of tags inidentification system. RFID tags can be divided into two aspects according to the range ofapplication: the one is the condition with a small number of tags such as small parking lot, smallsupermarket, production workshop, etc; the other is the condition with a large number of tagssuch as large livestock farm, comprehensive shopping mall, large logistics company, airport,large product supply chain, etc. This paper mainly studies the tag collision from the two aspectsabove. Above all, the existing algorithm principle with its available developing problems will beanalyzed associated with the research of the relationship of the characteristic of label distributionand related data structure in order to put forward the corresponding improvement scheme.To begin with, a tags anti-collision algorithm of dynamic frame timeslot based ondichotomy research will be proposed after the study of the occasion of small number of tags.Through the analysis of the traditional and typical dynamic frame timeslot tag anti-collisionalgorithm which committed mainly to study the frame length adjustment during the tagidentification process and the estimating of tags amount, we have found the fact that it is ignoredthat the characteristic of tags’ distribution, entered the reader’s recognition range, which isfollowing Poisson distribution. Therefore, this paper analyzes the Poisson distribution as well asthe difference and contact of tags’ distribution deeply. We can study the frequency of tags’occurrence after analyzing the slope characteristics of Poisson distribution. According that andcombining with the unique characteristics of data structure of dichotomy research, the relatedtheory could be applied appropriately to tags’ identification process of timeslot adjustment strategy to further improve the recognition efficiency and success rate of recognition. After thesimulation and experimental results analysis, the feasibility of the method can be verifiedthrough theory and actual data results.Furthermore, the anti-collision algorithm based on tags grouped DFSA samplingtraining rules has been proposed through the study of the occasion of large number of tags. Inthis case, the dynamic frame slot anti-collision algorithm encountered bottleneck. Therecognition efficiency and recognition rate fell sharply by the reason that the recognitionefficiency unable to load a large number of tag load. Therefore, it is necessary to group the tagsto be identified respectively. Many people dedicated to the study how to group tags to improvethe recognition efficiency and success rate. But the same grouping condition, roughly equalinterclass tag number and regularity of timeslot selected by tags make the frame lengthadjustment appearing repeatable fluctuations. Within the group tag identification process, it isalways neglected for reference value of regularity recorded history on frame length adjustment.This paper mainly studies the availability of historical record and applied it to the strategy offrame length adjustment. Then we can train it into rules to give a reference for the further tagsidentification. At last, the available of this method is verified by the simulation results. Theexperiment shows that this algorithm not only does not affect the success rate of recognit ion, butalso can improve the system recognition time, reduce the system computational complexity andsystem power consumption.