Verifying Bounded Subset-Closed Hyperproperties

Hyperproperties are quickly becoming very popular in the context of systems security, due to their expressive power. They differ from classic trace properties since they are represented by sets of sets of executions instead of sets of executions. This allows us, for instance, to capture information flow security specifications, which cannot be expressed as trace properties, namely as predicates over single executions. In this work, we reason about how it is possible to move standard abstract interpretation-based static analysis methods, designed for trace properties, towards the verification of hyperproperties. In particular, we focus on the verification of bounded subset-closed hyperproperties which are easier to verify than generic hyperproperties. It turns out that a lot of interesting specifications (e.g., Non-Interference) lie in this category.