论文部分内容阅读
设计了一种面向大规模嵌入式设备固件的自动化分析方法,该方法能够对固件进行自动化分析,提取其文件系统、操作系统、中央处理器指令架构等关键信息.针对固件解码成功的自动化判定难题,提出了一种基于分类回归树的固件解码状态检测算法,并选取收集的6 160个固件和固件自动化解码后得到的1 823个可反汇编二进制文件作为样本进行实验.实验结果表明,该算法相对其他分类器具有更好的分类效果,其分类准确率、召回率均在96%以上.
An automatic analysis method for large-scale embedded device firmware is designed, which can automate the analysis of the firmware and extract key information such as file system, operating system and CPU instruction structure.Aiming at the difficulty of automatic decision of firmware decoding success , An algorithm for detecting the status of firmware decoding based on classification and regression tree is proposed and 1823 disassembled binary files obtained by automatic decoding of the collected 6 160 firmware and firmware are selected as the experimental examples.The experimental results show that the algorithm Compared with other classifiers have better classification results, the classification accuracy, recall rate of 96% or more.