Study On Anti-collision Algorithm In RFID System

Radio Frequency Identification is a kind of non-touching automatic identifying technology which develops in recent years. RFID has many advantages compared with barcode, such as long reading distance, strong penetrating object, high efficiency, large data storage capacity and so on. RFID has a wide range of application in many fields. Therefore, RFID technology becomes the focus of researchers. Although RFID technology has many advantages, and in many areas has good prospects, there are still many problems to be solved. The problem of tag conflict is one of them. Some of the conclusions and contribution of this paper are listed as follows.A brief description of the RFID system is presented. Tag collision problem is analyzed in details. The advantages and disadvantages of existing anti-collision algorithm are analyzed. Current methods to solve this problem are Aloha algorithm and binary search algorithm. The identification of the binary search algorithm is more efficient. This algorithm exposes information and has poor security. Aloha algorithm consists of slotted Aloha algorithm, frame slotted Aloha algorithm and dynamic frame slotted Aloha algorithm. The efficiency is low when identifying the large number of tags.Through the analysis of Aloha algorithm, we find that the Aloha algorithm's recognition rate declines sharply with the increase of tags'number. This paper proposes an improved algorithm for dynamic frame slotted to make up with the weakness. The algorithm counts the number of collision slots in a frame, then estimates the number of tags according to collision slots and frame length,dynamically adjusts the frame length according to tags. Experiments show that the recognition efficiency of improved algorithm is better than minimum value algorithm.The efficiency of the methods declines sharply when the tags is greater than 500. The grouping algorithm can solve this problem. The algorithm estimates tags with statistical algorithm and then divides these tags into different group. It adjusts the frame length using minimum algorithm in each group. Experiments show that with the increase of tag number, the efficiency of grouping algorithm is stable around 0.34 and the efficiency of non-grouping algorithm decreased to 0.28 when the tag number is 1000. Apparently, the efficiency of grouping algorithm is higher in a large number of tags.