An environmental paradigm for defending security protocols

Abstract

In scenarios with multiple non-collaborating attackers, interference between simultaneous attack procedures can emerge. Interference has a wide range of consequences: it demands network models capable of supporting concurrence, it marks an unexpected complexity of the network environment, it can be exploited to construct partial defenses for vulnerable security protocols. In this paper, we describe a paradigm shift to support the investigation of interference with applications to constructing executable defenses for vulnerable protocols. We recruit tools from fields, such as AI and robotics, that have a suitably complex notion of agent, introduce a mechanism for the controlled execution of a protocol run and formalize it in terms of labeled transition systems. Finally, we show how the general problem of constructing defenses can be addressed systematically within the paradigm.