Comparative Analysis of Stateful Fuzzing Techniques Applied on EDHOC Implementations

Abstract

This thesis explores the concept of stateful fuzzing, a specialized form of fuzz testing that considers the state of the system being tested. It presents a novel approach where the underlying state machine model of the system under test is learned before the actual fuzzing process, rather than during it. The effectiveness of this method is compared with state-of-the-art fuzzers, AFLNet and StateAFL in the context of the EDHOC protocol, a secure key exchange protocol.