Verifying Security Policies using Host Attributes

Abstract

For the formal verification of a network security policy, it is crucial to express the verification goals. These formal goals, called security invariants, should be easy to express for the end user. Focusing on access control and information flow security strategies, this work discovers and proves universal insights about security invariants. This enables secure and convenient auto-completion of host attribute configurations. We demonstrate our results in a civil aviation scenario. All results are machine-verified with the Isabelle/HOL theorem prover.