A context-aware Multimodal Biometric Authentication for cloud-empowered systems

In the context of emerging technologies, Cloud Computing (CC) was introduced as a new paradigm to host and deliver Information Technology Services. In such an environment, privacy and security issues are critical areas that still require to be deeply explored. Yet, highly secured systems are generally met by computationally expensive systems. Such a system may also degrade the user experience and its willingness to adopt it. The aims of this paper are threefold: First, to integrate a Class-Association Rules (CARs) into the process of Multi-factor Authentication Based on Multimodal Biometrics (MFA-MB) for CC; Second, defines a new metric to measure the User Experience and; Third, exhibits an algorithm to authenticate cloud SaaS/PaaS Users with an enhanced MFA-MB scheme. Since, CARs are used to predict the most expected Multimodal Biometric Authentication (MBA) in the basis of Users' authentication habits, mined from their historical authentication data sets, which guarantees continuously improving their Experience. The integration of CARs in the CC authentication process allows also to identify the actual context (Time, Place, Device, etc.) which impacts the choice of biometrics used in MBA according to one User's situation. Therefore, this will help to increase the authentication security level using MBA at a decreased time and improved user experience. Integration of CARs is illustrated by a realistic case of Bimodal Biometric Authentication.