RFID signal acquisition and identification

Abstract

RFID is one of the fastest growing technologies grouped under Automatic Identification (auto ID). RFID tags are very low cost and used for identification of objects. RFID is a global technology that is used in industries, medical, wall mart, airport baggage, Libraries, Smart cards, even in every transported object has its own RFID tag. Therefore, concern of security and privacy should be there to prevent unauthorized access. A method is proposed to prevent cloning and counterfeiting of tags based on RF Fingerprinting. RF fingerprinting of a tag is based upon physical attributed such as an electromagnetic (EM) signal of the tag. By capturing the EM signal of RFID tags a method known as Dynamic wavelet fingerprinting is applied to generate fingerprint images of signals. Our proposed method consists of four stages: Namely Real time data acquiring by use of a CRO, Dynamic wavelet fingerprinting (DWFP) of the signal, Feature extraction, and Classification. Feature is extracted such as Eccentricity, perimeter, centroid , extent, area and orientation. Ann classifier is used which is a one vs. One classifier. To improve the performance of classification multi-feature based serial feature fusion technique has been proposed, which shows a significant improvement in classification performance. RF fingerprint allows prevention of unauthorized access, identification and detecting cloning of sensitive devices. To identify tags and to detect counterfeit RF fingerprinting can be used. The cost of the tag does not increase and can be used in existing tag with only requirement of software