On Dynamic & Subjective Trust for Privacy Policy Enforcement in Cloud Computing

Trust relationships are subjective and dynamic. One entity's trust over a subject is completely independent of some other entity's trust: trust is subjective. Trust also evolves dynamically over time. The growth of cloud computing over the last few years has created an environment where clients need to transact with cloud service providers even though the cloud service clients may or may not completely trust the provider at any given time. There is a need for the provider to acknowledge this dynamic and subjective trust assessment by each client by providing and modifying services individually for each client based on the trust assessment. This paper proposes a policy based approach to the implementation of subjective and dynamic trust so as to enable privacy policy enforcement in a cloud computing environment. We present an abstract model containing computational, storage and monitoring unit with configurable elements and describe algorithms that reflect how a change of trust influences the configuration of these elements. We prototype a system based on the abstract model with benchmarks.