Study On Parallel Analysis Method Of Internet Of Things Firmware Security

In recent years,the vigorous development of the Internet of Things(Io T)has brought convenience to people's daily lives,as well as severe security issues.In particular,the vulnerabilities hidden in the Io T firmware have become the new hotbed of cyber threats,such as botnets,ransomware,and advanced persistent threat.At present,the defense of Io T devices is facing many challenges.On one hand,it is difficult to deploy a defense-in-depth system on the resource-constrained Io T platform;on the other hand,the vulnerability mitigation measures for the Io T platform are not yet mature,resulting in classic vulnerabilities in the firmwares can easily become a breach for attackers.Therefore,it is necessary to study how to find vulnerabilities in Io T firmware as early as possible and repair them to deal with the threats brought by Io T security issues.The purpose of this paper is to analyze and discover the vulnerabilities of Io T firmware,starting from two representative technical routes of dynamic simulation execution and static taint analysis,to study the problems of automated analysis methods of Io T firmware security during batch operation,and improve the efficiency through parallelization methods.The main results of this paper are as follows:· This article analyzes the common errors of the dynamic simulation execution method when simulating the Io T firmwares,and designs a set of corresponding automated repair strategies.Then proposes a firmware running status detection method based on command line interaction.Experiments on a total of 32,834 firmware sample sets have verified that the success rate of the automatic repair strategy and detection method proposed in this article has reached 91.76%.The experiment also counts the distribution of firmware instruction set architecture,which provides a solid foundation for batch analysis of firmware status and working principles.· The current cutting-edge analysis method in the field of Io T firmware is multi-binary dataflow taint analysis method.It suffers from low efficiency and poor compatibility.This paper proposes a parallel taint analysis task scheduling method and conducts experiments.The experimental results show that the new method proposed in this paper is 9.6 times faster than the original method in terms of analysis speed,and the time to detect the same vulnerability is only 8.6% of the original method.In addition,this article also points out the process flaws of the original method when dealing with MIPS architecture firmwares,and fix the issue to improve the compatibility of the original method.