Study On Radio Frequency Identification Anti-collision Algorithm

With the bringing up of Internet of Things, various countries rapidly expand thefierce rivalry in the field of information technology. As one of the core technology ofthe Internet of Things, RFID systems get rapid development in recent years. RFID’sbiggest advantage is the ability to multi-target recognition, which is also the technicalkey urgent to overcome. The collision problem in the field of multi-target identificationhas a very negative impact on the identification of tags in RFID systems. Therefore,how to establish an efficient anti-collision algorithm is a research topic badly needed tobe resolved both at home and abroad.In allusion with problems of tag collision and reader collision in the process ofmulti-target recognition in RFID system, the two classical types of tag anti-collisionalgorithms are respectively optimized in this paper, the names of which are theuncertain tag anti-collision algorithm based on ALOHA and the deterministic taganti-collision algorithm based on binary tree. Meanwhile, the reader anti-collisionalgorithm is analyzed and an improved algorithm is proposed in this paper. The mainwork done in this dissertation is as follows:(1) Firstly, research on the composition of RFID system is made. Analysis of theworking principle of RFID system and the basic working processes as well as the wayof energy transmission are studied in this paper.(2) Due to the limit of the maximum number of digits in the ISO18000-6standard,a solution is proposed which adapts to the condition that the number of tags is largerthan the maximum of frame length. The improved algorithm uses the idea of groupingand a mechanism of frame length adjustment, increasing the system throughput throughlimiting the number of tag to respond. Tags are grouped in this algorithm firstly and ineach group the frame is adjusted dynamically. After tags of all slots in a frame areidentified, the number of remaining tags in this group is estimated. The algorithm is animprovement on uncertain tag anti-collision algorithm based on ALOHA.(3) The tag anti-collision algorithm based on binary tree is elaborated minutely inalgorithm ideas, algorithm steps and performance analysis, and some examples areexerted to discuss. Due to fact that the objects to be identified in the multi-targetidentification applications are usually similar to each other, an improved algorithmwhich is more close to the actual situation by combining with the backward indexing algorithm and the jumping dynamic tree anti-collision algorithm is proposed. Theimproved algorithm deals with high continuous collided bits separately, using the stackto store the high collided bits of EPC. As a result, both the system throughput and thespace complexity are improved, as well as the reader-side storage cost is reduced.(4) The reason why collision of readers comes about is explained, and the existingmulti-reader anti-collision algorithms are discussed in detail. In order to reduce thecomplexity of the hardware design of the reader and to get rid of the central nodecontrol so that it is more suitable for the frequent changes of the topological structure ofthe mobile reader network, an improved algorithm is proposed with experimentalsimulation conducted to evaluate its performance.