Research On Anti-collision Algorithm In RFID System

As one of the key technologies of Internet of things, RFID technology will getlarger and larger development space with the guide of Internet of things.Accomplishing automatic identification of multiple far field contactless objects, RFIDcan be applied in all aspects of daily life such as traffic, military, industrial and medicalaffairs. However, RFID is confronted with information collision due to contactlessapplications. The information collision limits the application of RFID, bringing manyreading issues on mistakes and omissions. Consequently, collision issue must be solvedcarefully during the development of RFID technology.The thesis introduces the structure of RFID system and work principle firstly.Focusing on RFID’s structure, the inner component and function of Reader and Tag aredemonstrated; focusing on work principle, the process of sign coupling, modulating,coding and so on are described. Furthermore, the thesis points out that the anti-collisionalgorithm is very important in RFID technology, which plays an important role in dataintegrity aspect.Then the thesis researches the tag anti-collision strategies of RFID system, whichincludes two aspects, probabilistic anti-collision algorithms based on ALOHA anddetermined anti-collision algorithms based on binary tree. The anti-collision algorithmsbased on ALOHA mainly includes the evolution from pure ALOHA, slotted ALOHA,frame slotted ALOHA, to dynamic frame slotted ALOHA (DFSA). The latter algorithmmade better performance on the number of collisions or system efficiency than theformer one. The anti-collision algorithms based on binary tree began with binarysearching, and then dynamic binary searching due to information redundancy, thencapturing backward searching owing to the number of searching, afterward, binarybound searching algorithm on account of information redundancy and the number ofsearching. Meanwhile, the implementation procedures on several algorithms areanalyzed and the performances are compared.At the end of the thesis, an optimal timeslot strategy for DFSA anti-collisionalgorithm is proposed, which considers the influence of the idle and the collisiontimeslots on system efficiency. In the practical situation, we short the idle and the collision timeslots, get the optimal parameter based on maximum throughput, anddetermine the dynamic frame length by optimal parameter and unread tags. Theproposed algorithm performs tag estimation with optimal Chebyshev’s Inequality bycomparing the accuracy rate, adopts Markovian to analyze the process of identification,controlling the read cycles effectively. Finally, the performances of the novel algorithmon accuracy rate of tag estimation, system efficiency and average identification time areverified with matlab. The simulations indicate that the proposed DFSA algorithm hasthe higher accuracy and stability in tag estimation, the better system efficiency, and theshorter identification time than the original algorithms.