Skip to main content


Formal verification tools

Project overview

Most weaknesses in security protocols are not due to flaws of the underlying cryptography, but they are caused by logical mistakes in the protocols’ construction. The detection of these mistakes is particularly difficult, since we are concerned with a distributed system where some nodes and the communication medium are hostile. The human mind can easily overlook possible shortcomings in such a setting. We therefore need automated methods to rigorously analyze security protocols and security-critical systems.

Formal verification tools


IBM supports the development of the OFMC tool, the On-the-Fly Model-Checker for security protocols. As shown in the figure (right), the user specifies the protocol and the goal (e.g. mutual authentication) as the input for OFMC:

As is standard in formal protocol verification, OFMC uses an abstract model of cryptographic primitives (such as symmetric and asymmetric encryption, hash functions, or modular exponentiation). In contrast to most existing approaches, this interpretation is not fixed, but can be chosen by the user in form of an algebraic theory of the used function symbols. As a result, OFMC either replies with “Verified” in case no attack is found or with an “Attack Trace” which illustrates a violation of the goal that can happen.

back to top

Publications

back to top

Images

CLESS

click to enlarge Figure 1. OFMC tool.