A Robust Biometric Authentication and PIN Distribution Technique for Secure Mobile Commerce Applications

摘要

In a mobile emerging world, user authentication, service provider authentication and security is very important in mobile commerce. User authentication is performed by using fingerprint based biometric methodology. Existing system used for Mobile purchasing/payment services in handheld devices does not analyze fingerprint matching and feature extraction techniques in an efficient way. Also the existing system is not secure and accurate for m-payments applications. We propose secure, efficient and accurate m-commerce architecture for m-commerce applications. This involves fusion of Minutiae Maps (MM) and Orientation Maps (OM) for fingerprint feature extraction. The fingerprint is sent to the biometric server in a secure way using Discrete Wavelet Transform (DWT) data hiding method. User fingerprint will be checked and compared using MM and OM methods to figure out the fingerprint threshold matching score. If the threshold is 80-99% PIN distribution process is initiated, otherwise user authentication is failed. The user PIN is converted into a unique sequence and divided into two parts. Along with the user PIN, user IP address, time stamp, user ID are encrypted using RC4 (stream cipher) algorithm. Also a hash function is appended to the cipher text using Secure Hash Algorithm (SHA4). One half is verified by the authentication server and the other half is verified by the external server. After verification both the servers sends only OK message to the bank. This study looks to provide a high secure and efficient solution for m-commerce applications.