A Cryptographic Protocol Vulnerability Analysis Framework based on Fuzz Testing and Model Learning

Aiming at the increasingly prominent security issues in cryptographic protocols, based on the model learning technology in automata theory, the black-box test of cryptographic protocols is carried out by using the protocol state fuzzy testing technology, and a cryptographic protocol analysis system is designed and implemented to realize inferencing the target protocol's state machine automatically. The proposed system is used to test the OpenVPN protocol, clearly showing the specific behavior characteristics of the protocol and multiple special behavior paths outside the expected behavior path and possible security analysis and comparison, which has important reference significance on the vulnerability analysis of other cryptographic protocols.