Vulcloud: Scalable and Hybrid Vulnerability Detection in Cloud Computing

Vulnerability exploits will result in security breaches or violations of the system´s security policy causing information leakage or economic losses. Although many detection methods such as static analysis, dynamic analysis and fuzz testing have been presented, the vulnerabilities are still difficult to detect. In this paper, we propose a new detection cloud service Vulcloud, which is scalable and hybrid combining the static, dynamic and fuzzing into cloud computing. Vulcloud first statically analyzes the objects and reports the potential vulnerable items. And then, the fuzzing cases for the items are semi-automated created, and tested under the dynamic monitoring. Finally, the source code of the results are statically analyzed again to determine whether they are vulnerabilities or not. The prototype of Vulcloud is implemented, and the performance is evaluated by Mplayer source code. The experiment results show that Vulcloud can detect vulnerabilities in software, and the challenges of storage and processing capabilities are resolved by cloud computing.