Cognitive Risk Control for Physical-Layer RFID Counterfeit Tag Identification

Physical-layer counterfeit tag identification is a widely adopted technology for detecting clone tags in radio-frequency identification (RFID) systems. It has gained significant attention in recent years owing to its affordability and ease of implementation. However, the effectiveness of physical-layer identification technology is highly dependent on the tag identification environment. When the reading distance is far or signal power is low, the signal-to-noise ratio (SNR) of the tag signal may be low, leading to a degradation in classification accuracy. To address the issue, this article explores the application of cognitive risk control (CRC) in tag classification, aiming to enhance the SNR and improve classification accuracy. Additionally, this article introduces more time- and frequency-domain features, resulting in a total of 104 features, to further enhance the classification performance. The experiment utilizes a software radio device to classify tags from three manufacturers across seven popular types. The results indicate that the implementation of CRC has, on average, improved the classification accuracy of the tags by approximately eight percentage points. Moreover, when compared to traditional 28 and seven features, the use of 104 features has shown an average increase in classification accuracy of about two and four percentage points, respectively.