Abstract

The fast growing popularity of cloud computing within enterprise and critical infrastructure settings has increased the need to develop secure, adaptive and intelligent mechanisms of access control. The traditional and more static and role-deterministic approaches are becoming less useful in managing dynamically variable and contextual security needs. This paper develops an AI-enhanced secure access control model which incorporates both context-aware characteristics and ensemble learning to improve cloud security. The system is able to adaptively analyse user actions and the context of the environment and the sensitivity of the resources through a system of machine learning classifiers in an ensemble methodology, providing high accuracy and noise tolerance. Based on experimental verifications, the model suggests it will be useful in cloud access governance due to its better performance in detecting malicious servers, a low latency, and high resistance to adversarial threats than conventional approaches, which provides a solid base to scale and resilience cloud access governance.

References

Download PDF
Back