Using edit automata for rewriting-based security enforcement

Abstract

Execution monitoring (EM) is a widely adopted class of security mechanisms. EM-enforceable security properties are usually characterized by security automata and their derivatives. However Edit automata (EA) have been recently proposed to specify more powerful EMs. Being able to feign the execution of sensitive program actions, these EMs are supposed to enforce more security properties. However, feigning program actions will usually make the program behaving in discordance with its specification since the effects of feigned actions are not reflected in the program states. In this paper we highlight this problem and show how program rewriting can be a reliable enforcement alternative. The paper contribution is mainly a semantics foundation for program rewriting enforcement of EA-enforceable security properties. © 2009 Springer Berlin Heidelberg.