Research Of Multi-tags Anti-Collision Algorithms In Passive RFID System

Radio Frequency Identification (RPID) technology, uses radio frequency signal as the medium, and through space coupling technique, to achieve the goal of non-contact identification. As one of automatic identification technologies, RPID technology has the unparalleled advantage of other similar technologies, such as rewritable, environmental adaptability, long distance identification, can "simultaneously" identify several tags and can carry large volumes of data. Therefore, RFID technology will certainly get more and more widely used.But there are many obstacles in the way of its promotion process, the tag collision problem is one of the key factors. The tag anti-collision mechanism is essential since well-designed multi-tag anti-collision algorithms can significantly improve the recognition efficiency of the RFID system.This paper makes some analysis based on multi-tag anti-collision algorithms for passive RFID systems, and then makes some improvements, in order to achieve a better performance.Firstly, some analysis is made based on the Binary Search Algorithm and its improved algorithms, and then some improvements is made based on the Backward Dynamic Binary Search(BDBS) algorithm:1) By using the method of shielding non-collision bits, the average amount of data sent during one query process is greatly reduced, so as to achieve the purpose of performance improvements and reducing energy consumption;2) Through the application of the pre-order hybrid query tree, the system performance is improved since the reduction of the number of queries required, the simulation results show that the system performance have been greatly raised, but still has the potential of improvement.Secondly, this paper make some analysis of the ALOHA random anti-collision algorithms, after that some improvements is made based on the Dynamic Framed Slotted ALOHA(DFSA) algorithm:that the collision caused by two tags responding at the same time is solvable, both of them are able to be identified, as a result, the case when two tags arrive at the same slot should not be considered unable to identify. Calculations show that the optimal frame size reduced by nearly30percent, that is, if the frame size is certain, the number of tags that can be handled increased significantly; and on the basis of the Poisson distribution theory, a new tag estimation formula can be obtained; Finally, calculations show that the improved algorithm make a significant increase of the system throughput indicator.