A blended approach of static binary mining and exploratory data analysis to obtain the security posture of embedded systems firmware

In the era of connected embedded systems, devices are often targeted by sophisticated cyber-attacks. The entry point for these attacks is more often through firmware and hence analysing its security is vital. This research presents an innovative method for evaluating the security posture of firmware by examining its binary constituents. The approach combines static binary mining techniques with exploratory data analysis to visually depict the security posture of firmware. This aids in providing a comprehensive perspective of its weaknesses and strengths. To validate the effectiveness of the proposed approach and methodology, an in-depth security analysis is carried out on five real world firmware samples. The selection of these firmware samples encompasses a wide range of devices and applications, thereby assuring the broad relevance of the findings. The results demonstrate the efficacy of this technique in revealing and visually representing different firmware security characteristics, such as the firmware attack surface, the effectiveness of binary protection mechanisms, and the overall security resilience. The methodology is implemented as an open source tool named fw2ai, which automates the security analysis serving the needs of both security researchers and practitioners.