| With the popularization of IoT devices,the attacks against IoT devices have emerged in large numbers in recent years.Among the attacks against IoT devices,command injection attack is one of the common and effective attacks which injects commands into the target program,and further obtain the control authority of the target device.However,in the field of Internet of Things security research,there are few researches on its detection and certain limitations exist.For example,the dynamic analysis fuzzing technology mainly focuses on memory corruption vulnerabilities,and is subject to effective input while static analysis symbol execution technology is subject to heavy execution overheads or false positives and false negatives leading to low detection efficiency.This paper proposes an intelligent dynamic detection technology for command injection vulnerabilities of IoT devices to make up for the shortcomings of current research on the detection of command injection vulnerability of IoT devices.The work of this paper are as follows:1)Proposing an intelligent dynamic detection model for firmware command injection vulnerabilities,which introduces the front-end file logic analysis,the program information extraction technology based on static analysis and a fuzzing method based on the feedback from distance function,improving the efficiency and accuracy of detection.2)Proposing a sample generation technology based on the front-end file logic analysis,which analyzes the interaction code between the front-end and the back-end and generates samples that conform to the back-end program,enhancing the penetration capability of samples.3)Proposing a program information extraction technology based on static analysis which screens tainted functions,filters key interface functions,and sets probe codes at corresponding positions to monitor the execution of samples.4)Proposing a fuzzing method based on the feedback from Distance Function which selects high-quality samples and key fields for mutation by evaluating the performance of samples to improve the efficiency of command injection vulnerability detection.Through testing and experimental comparison,compared with SaTC and Commix,the intelligent dynamic detection system IoTCID proposed in this paper for command injection vulnerability of IoT devices has better performance in the generation of samples and the detection of command injection vulnerabilities.IoTCID confirms multiple public and 2 undisclosed command injection vulnerabilities.It shows that IoTCID are effective in discovering command injection vulnerabilities in IoT devices. |
