- Gates&Hillman Centers
- Traffic21 6501
- ARBOB AHMAD
- Computer Science Department
An Epistemic Formulation of Information Flow Security
The non-interference (NI) property defines a program to be secure if changes to high-security inputs cannot alter the values of low-security outputs. NI indirectly states the epistemic property that no low-security principal acquires knowledge of high-security data. We consider a directly epistemic account of information flow (IF) security focusing on the knowledge flows engendered by the program's execution. Storage effects are of primary interest, since principals acquire knowledge from the execution only through these effects. The IF properties of the individual effectful actions are characterized using a substructural epistemic logic that accounts for the knowledge transferred through their execution. We prove that a low-security principal never acquires knowledge of a high-security input by executing a well-typed program.
The epistemic approach has several advantages over NI. First, it directly accounts for the knowledge flow engendered by a program. Second, in contrast to the bimodal NI property, the epistemic approach accounts for authorized declassification. We prove that a low-security principal acquires knowledge of a high-security input only if it is authorized by a proof in authorization logic. Third, the explicit formulation of IF properties as an epistemic theory provides a crisp treatment of ``side channels.'' Rather than prove that a principal does not know a secret, we instead prove that it is not provable that the principal knows that secret. The latter statement characterizes the ``minimal model,'' for which a precise statement may be made, whereas the former applies to ``any model,'' including those with ``side channels'' that violate the model's basic premises. Fourth, the NI property is re-positioned as providing an adequacy proof of the epistemic theory of effects, ensuring that the logical theory corresponds to the actual program behavior. In this way we obtain a generalization of the classical approach to IF security that extends to authorized declassification.
Robert Harper (Chair)
Aleksandar Nanevski (IMDEA Software Institute)