Evolving Authentication Design Consideration and BaaS Architecture for Internet of Biometric things

Abstract

Cloud computing has opened many possibilities for the design and implementing various models of software for addressing variety of problems. Biometric authentication is an important aspect of security and need for the same is increasing day by day. This has resulted in adaptation of biometric security-based systems by masses. Further, with the portable sensors and computing devices the biometric authentication will become an integral part of Internet of Things. To handle such a large scale of authentication cloud-based implementation is a viable option. In this paper, an evolution of a multimodal biometric authentication system is discussed. Initially the system is proposed as a native multimodal biometric system and further modified to be deployed as Software as a service-based model and a cloud-based model with GPU and NoSQL database. This paper is presenting the improvement in computation time and scalability gained by cloud-based deployment.