Fuzzing Improving Techniques Applied and Evaluated on a Network Traffic Analysis System

With the rapid advancement of modern technology, software are getting more and more intelligent, whereas the related security issues are also becoming more and more prominent [1]. Developers can easily introduce bugs into a system due to mistakes or poor consideration, which in serious cases may lead to security breaches that can jeopardize the entire system or the entire network. Therefore, efficient and automated security testing techniques or theories are of great importance to software testing and vulnerability detection. Fuzzing is currently one of the most popular vulnerability discovery [2] techniques, not only because of its reliability and numerous successful cases, but also because of its unlimited potentials and possibilities. Fuzzing may not be able to find find vulnerabilities on its own, but it certainly can lead us to faster and more efficient vulnerability discoveries. In this paper, we'll fuzz a Network Traffic Analysis (NTA) System. We discuss some of the difficulties we meet in fuzzing the system and apply appropriate fuzzing techniques to tackle those difficulties. Finally, the fuzzing techniques applied are evaluated, and the applicability and characteristics of these techniques are summarized.