Survey of network protocol fuzzers: Taxonomy, techniques, and directions

Abstract

Fuzzing has become widely adopted for network protocol vulnerability detection due to its high degree of automation and minimal reliance on domain-specific knowledge. Given the distinct characteristics of network protocol programs compared to general targets, researchers have proposed numerous innovative solutions to address the technical challenges. However, there remains a lack of thorough investigation that provide in-depth technical analysis and comprehensive summarization of these advancements, as well as a clear taxonomy to guide future research directions. To bridge this gap, this study conducts a systematic review of network protocol fuzzing and proposes a novel framework with four core modules abstracted from protocol fuzzer architectures. We analyze the key technologies in each module, discussing their advantages, limitations, and application scenarios. More significantly, this work establishes a novel taxonomy defining four fundamental capability dimensions, each addressing distinct practical challenges in protocol fuzzing. Using this framework, we conduct the systematic classification and comparative analysis of existing techniques. Our work contributes theoretical insights and practical guidance for network protocol fuzzing development.