Review of Firmware Homology Detection: a System Aspect

Review of Firmware Homology Detection: a System Aspect

연구 분야: Analysis

학회: 2021 IEEE Sixth International Conference on Data Science in Cyberspace (DSC)

In recent years, Internet of Things (IoT) devices have received extensive attention and can be seen everywhere in daily life. Attacks against firmware vulnerabilities are becoming more and more frequent. Homology analysis technology is a hot topic in recent years. Firmware vulnerability analysis based on homology analysis can classify the firmware of different versions or even different devices, and detect the vulnerabilities in the firmware by detecting the homology of the open source library. This paper provides a literature review of firmware homology detection. Firstly, the background of firmware vulnerability analysis and the research status of homology detection at home and abroad are briefly introduced. Then, under the source code based homology analysis technology, according to whether it chooses to use the intermediate form to explain the source code and the different intermediate forms, three different homology analysis technologies based on text, token and structure are divided, and the three technologies are compared horizontally. Then, in the non source code based homology analysis technology, the overall process of homology detection based on binary files is introduced, as well as the main technologies such as byte stream comparison and reverse analysis. Finally, various homology analysis methods are analyzed and compared from the aspects of efficiency and accuracy, and the research focus and direction in this field are summarized and prospected.