Toward Firmware-Type Analysis Using Machine Learning Techniques

摘要

Due to the development of the Internet, a considerable amount of firmware that operates on various types of hardware has been developed. The process of analyzing firmware according to these developments has also been important. Among the various tools for analyzing firmware, the tool for analyzing firmware types is important as the first step of the process of analyzing firmware. This paper introduces a signature-based firmware analysis approach of the type mainly used at present and proposes a new idea for analyzing firmware types based on machine learning.