A Multimodal Biometric Authentication (MBA) in Card Payment Systems

Abstract

The authentication method that is most commonly used in bank card payment systems is a Personal Identification Number (PIN). This method of authentication has many drawbacks, one of them being that the card-holder has to remember the PIN from a smart card or from many smart cards, and the other drawback is the possibility that the PIN data may be damaged. In order to address these shortcomings, a few models have been developed to create a PIN from biometric data. The idea for the development of the algorithm proposed in this paper came about through the research of similar works such as "Generation of Reliable PINs from Fingerprints," Fengling Han at al., with the aim of making the algorithm more stable and creating the possibility of authentication without the need for the authentication system to store the personal biometric data of the cardholder. The purpose of this paper is to address the process of entering the PIN easily, by tapping your finger on a fingerprint scanner and taking a photo of the card-holder's face. The algorithm presented in this paper encodes multimodal biometric fingerprint data and interpupillary distances of the left and right eyes into a 4-digit or 6-digit PIN, at the same time, as the process of its verification takes place. Thus, the encoded PIN is compared to the PIN placed on the bank card's chip. The advantage of this multimodal biometric authentication (MBA) model is that it does not require comparison of biometric data with data in a bank's authorization center. The results of experimental studies confirm that using the presented MBA model, the cardholder does not need to remember PIN values to make card payment transactions, and therefore the card-holder's user experience is enhanced.