More about HKUST
Fuzzing: A Survey and Reflection
PhD Qualifying Examination Title: "Fuzzing: A Survey and Reflection" by Mr. Heqing HUANG Abstract: Fuzzing is known to be one of the most efficient methods to detect real-world vulnerabilities. It repeatably running the target programs with enormous malformed inputs generated automatically. Unlike heavyweight analysis, fuzzing requires less knowledge of the targets so that it can be easily scaled up to for large-size real-world programs. Its effectiveness has also been proved in detecting server vulnerabilities such as Heartbleed. Nowadays, there are lots of studies related to improving the fuzzing technique. Moreover, the increasing software's scale and usages are urged for ensuring its qualities. Therefore, to help understand the methodology of the state-of-the-art fuzzing techniques and its vast applications, we conduct a thorough survey related to the existing literature and details comparison among the different techniques. Since fuzzing aims to detect the program flaws, the key problem is to improve the quality of the generated inputs. Therefore, we first introduce the conventional input generation approach. Afterward, we divide the existing solutions into two categories. The first category relates to the sophisticated input generation method. With different optimization, fuzzing can generate inputs both efficiently and effectively. The second category focuses on prioritizing the inputs for fuzzing. This type of method makes fuzzing capable to tackle the path explosion problem. After conducting a thorough survey of the existing researches, we also provide advice for potential future directions. We believe our survey will shed light on our future work on fuzzing. Date: Tuesday, 14 April 2020 Time: 10:00am - 12:00noon Zoom Meeting: https://hkust.zoom.us/j/4348622853 Committee Members: Dr. Charles Zhang (Supervisor) Prof. Shing-Chi Cheung (Chairperson) Dr. Dimitris Papadopoulos Dr. Shuai Wang **** ALL are Welcome ****