作者: William E. Cobb , Eric D. Laspe , Rusty O. Baldwin , Michael A. Temple , Yong C. Kim
DOI: 10.1109/TIFS.2011.2160170
关键词: Computer science 、 Pattern recognition 、 Identification (information) 、 Artificial intelligence 、 Fingerprint 、 Multiple discriminant analysis 、 Multi-factor authentication 、 Cryptographic protocol 、 Authentication 、 Physical layer 、 Speech recognition 、 Access control
摘要: Radio-frequency distinct native attribute (RF-DNA) fingerprinting is adapted as a physical-layer technique to improve the security of integrated circuit (IC)-based multifactor authentication systems. Device recognition tasks (both identification and verification) are accomplished by passively monitoring exploiting intrinsic features an IC's unintentional RF emissions without requiring any modification device being analyzed. discrimination achieved using RF-DNA fingerprints comprised higher order statistical based on instantaneous amplitude, phase, frequency responses executes sequence operations. The system trained multiple discriminant analysis reduce data dimensionality while retaining class separability, resultant classified linear Bayesian classifier. Demonstrated verification performance includes average accuracy greater than 99.5% equal error rates less 0.05% for 40 near-identical devices. Depending level required classification accuracy, fingerprint-based well-suited implementation countermeasure cloning, promising use in wide variety related problems.