Information flow in a distributed security setting

TL;DR

This paper explores how information flow security can be maintained in distributed systems with multiple local policies, introducing new security properties and enforcement mechanisms for programs with location-dependent behaviors.

Contribution

It introduces Distributed Non-disclosure and Flow Policy Confinement properties, along with enforcement techniques, to ensure compliance with local and allowed flow policies in distributed settings.

Findings

Type and effect systems effectively enforce security properties.

Hybrid static-dynamic mechanisms improve enforcement accuracy.

Distributed Non-Interference guarantees policy compliance in distributed systems.

Abstract

Information flow security is classically formulated in terms of the absence of illegal information flows, with respect to a security setting consisting of a single flow policy that specifies what information flows should be permitted in the system. In this paper we investigate the security issues that emerge in distributed security settings, where each computation domain establishes its own local security policy, and where programs may exhibit location-dependent behavior. In particular, we study the interplay between two distinct flow policy layers: the declared flow policy, established by the program itself, and the allowed flow policy, established externally to the program by each computation domain. We refine two security properties that articulate how the behaviors of programs comply to the respective flow policies: Distributed Non-disclosure, for enabling programs to declare locally scoped flow policies; and Flow Policy Confinement, for controlling the flow policies that are declared by programs. We present enforcement mechanisms that are based on type and effect systems, ranging from purely static mechanisms to hybrid combinations with dynamic migration control, for enforcing the above properties on an expressive ML-like language with concurrent threads and code migration, and which includes an allowed flow policy construct that dynamically tests the allowed flow policy of the current context. Finally, we show that the combination of the above two properties guarantees that actual information flows do not violate the relevant allowed flow policies. To this end we propose and use the Distributed Non-Interference property, a natural generalization of Non-Interference to a distributed security setting that ensures that information flows in a program respect the allowed flow policy of the domains where they originate.