Achieve Fuzzing Based on Symbolic Execution Platform

Abstract

The traditional software development does not consider the concerns, which is can addressing security problem effectively. So how to find the problem is the standard consideration for security researchers. In the field of software security testing, there are four ways, Fuzzing, Dynamic Taint Analysis, Model Checking and Symbolic Execution. This paper we use symbolic execution to achieve functional of fuzzing. Introduction With the development of computer technology, there are more and more software applied in every walk of life, so the security of software have been known by developer. However, we cannot avoid the exception in developing software, the method of software vulnerability analysis is an important way to protect the security of information system. There are four ways in this field. Dynamic taint analysis [1] is belongs to data-path analysis. It sign the data with untrusted source to the taint, and tracking the data. Model checking [2] will model a system and uses formal method to judge the system whether it is a satisfied system properties. Fuzzing [3,4] can provide random input data to find the software vulnerability. Symbolic execution [5] is also belongs to data-path analysis, it use symbolic input instead of actual input, so it can generate more targeted test cases automatically.