Dependability Quantification of Cloud-Centric Authentication Frameworks

摘要

Password-based Authentication System (PAS) is one of the preferable solutions to access secure cloud services via Internet applications. In this regard, four well-known PAS systems viz; single server, multiserver, gateway augmented multiserver and two-server based PASs have been evolved in the existing state of the art. To assess dependable user authentication services in the cloud, high server-side availability of the aforesaid systems are necessary. In this paper, we propose a unified mathematical model for both single server and two-server based PAS systems. We put forward a Generalized Stochastic Petri Nets (GSPNs) based system modeling and Markovian state-based dependability analysis technique to quantify two important dependability metrics viz; reliability and availability for the said PAS systems. The performance analysis substantiates that the two-server based PAS is more dependable than the single server-based PAS in terms of availability. The proposed approach can be useful in the context of emerging cloud computing paradigm for measuring the security service dependability.