Study On The Identification Of Chip-less RFID Tags Based On Artificial Neural Network

Electronic tag is an important component of the radio frequency identification system. As a data carrier, it is widely used in many respects of goods tracking, identity recognition, and information gathering, etc. The cost of Chip-less tags is lower than that of tag with chip due to the absence of the microcontroller and on-board battery source. Meanwhile, for the chip-less tags, the relationship between the performance parameters and the scattered field is complex and nonlinear. It is very difficult to reveal the intrinsic law using the traditional methods. Therefore, it is very important to search an easy way to solve the identification problem. In this thesis, an identifying method based on neural network for chipless RFID tags is studied and the identification results is presented. Compared with traditional identification methods, the Identification of chip-less RFID tags base on artificial neural network is faster. This paper is organized as follows.Firstly, the research status of the neural network is presented. Then the basic theory, perceprton, the back-propagation algorithm (BP algorithm), and the basic characterizing parameters of the neural network are introduced.Secondly, study the scattering field characteristics of linear chip-less tag and establish the network model of the identification through analyzing the sample data of from CST. The identification of tag is realized by the network mode. Simulation results show that the identification error is less then 1.5 degree. Thus, when one encodes the data using angle, angular interval of 5° can be selected. In this case, all the tag can be identified accurately.Then, study the scattering field characteristics of V-shaped chipless tag and analyze the sample data of from CST. The network model of the identification system is established through training and amending. Simulation results show that the identification error is less then 5°. Thus, when one encodes the data using angle, angular interval of 10° can be selected.At last, study the effect of packaging material, such as paper, on the scattered field of V-shaped chipless tag. The identification results for different thickness are presented. Simulation results show when the material thickness is less than 0.2 mm, tags can be identified accurately.