Magnetic Stripe Authentication System for Handheld Devices

In this paper, we propose a portable magnetic stripe authentication system for banknote authentication by using the renminbi (RMB) magnetic stripe signal. The system uses a tunnel magnetoresistance (TMR) sensor to achieve high sensitivity and large dynamic range of signal acquisition. With the signal processing circuit, noise suppression and amplitude adjustment are realized. According to the statistical characteristics of the signal, three different kinds of denomination classification algorithms are implemented, based on the peak spacing feature, BP neural network and support vector machine, respectively. Experiment results show that these three algorithms have advantages in different application scenarios with high reliability.