Estado del arte utilizando mapeo sistemático para las técnicas de análisis de malware en android

Abstract

Malware attacks go to Android are a global prob lem, so that to prevent attacks on mobile devices, some techniques have been proposed to study the differents types of malicious programs. In this research, a state of the art was formulated using a systematic mapping, to extract the most relevant journals and conferences, from 2017 to 2021. The systematic literature review analyzed the Malware techniques in Android, in order to build a scheme of classification and put them into the structure of static, dynamic, or hybrid analyzes. The taxonomy revealed the methods, tools and learning systems more used, so that found the improvements and limitations of each of the analysis techniques. Quantitative data showed by the accuracy metric, were: static analysis = 95.36%, dynamic analysis = 92.44% and hybrid analysis = 96.81%. Therefore, these results show that the hybrid analysis has some advantages over the static or dynamic analysis technique, reducing its limitations and improving the detection of Malware in Android. Finally, the better option to analize malware is the supervised learning to classify the new generations of Malware from two features: their families and frequency used.