Abstract:
Cloud computing has been one of the major emerging technologies in recent
years. However, cloud computing presents an added level of risk because essential
services are often outsourced to a third party, which makes it harder to maintain data
security and privacy, support data and service availability, and demonstrate
compliance. Moreover, cloud computing comprises of various technologies like
virtualization, transaction management etc., so it also inherits their security issues.
The cloud computing technology introduces new security risks that need to be
assessed and mitigated. However, a traditional security risk assessment methodology
is not suitable to cloud computing due to its several characteristics. Recently, several
risk assessment methods and models have been proposed to assess the security risk in
cloud computing. None of these methods is fully quantitative. Moreover, none of
them are scenarios based to fit the dynamic nature of the cloud computing
environment. Therefore, assessing the security risk in cloud computing is still an
open research issue.
In this thesis we present a scenario-based methodology to assess security risk
in cloud computing. This methodology enables the provider to assess the security
risk in cloud computing applications. This methodology is based on the National
Institute of Standards and Technology (NIST) Risk Management Framework. In this
framework the risk is derived by multiplying the ratings assigned for threat
likelihood and the threat impact. We propose using Bayesian networks to determine
the likelihood which enables us to compute the probability of failures over variables
of interest given the evidence for the certain scenario of usage for the application. In
addition, we propose two methods to specify the impact factor. The first is to
categorize impact by expert assessment according to MIL-STD-882E standard
severity categories. The second method is using the worst case sensitivity analysis to
assess the threat impact.
To validate the proposed methodology we use two case studies, the Ecommerce
application, and a Live VM Migration scenario. As we compare the
proposed method with the existing methods base on assessing risk in the dynamic
scenarios. Furthermore, we apply security controls on a case study and the result
show significant reduction in risk values and mitigation for significant risk.
