Technique for IoT malware detection based on control flow graph analysis

The Internet of Things (IoT) refers to the millions devices around world that are connected Internet. Insecure IoT designed without proper security features targets many threats. rapid integration into infrastructure in various areas human activity, including vulnerable critical infrastructure, makes detection malware increasingly important. Annual reports from cybersecurity companies and antivirus software vendors show an increase attacks targeting infrastructure. This demonstrates failure modern methods for detecting on things. is why there urgent need new approaches protect attacks. subject research process Things. study aims develop a technique based control flow graph analysis. Results. paper presents approach Control graphs were built suspicious applications. represented as directed graph, which contains information about components program transitions between them. Based metrics can be extracted describe structure program. Considering applications small due simplicity limitations operating system environment, analysis seems possible environment. To analyze behavior application each action built. It shows abstract description application, sequence formed. allows defining program’s behavior. Thus, with aim detection, two models sequences used. Since you both overall it achieve high accuracy. proposed unknown malware, modified versions known malware. As mean conclusion-making concerning presence, set machine learning classifiers was employed. experimental results demonstrated accuracy detection. Conclusions. A has been developed. detect efficiency.