Towards Quantifying the (In)Security of Networked Systems

Traditional security analyses are often geared towards cryptographic primitives or protocols. Although such analyses are absolutely necessary, they do not provide much insight for answering an equally important question: what is the security assurance of a physically or logically networked system when we consider it as a whole? This question is known to be notoriously difficult, and the state-of-the-art is that we know very little about it. In this paper, we make a step towards resolving it with a new modeling approach.