Malware Detection with Obfuscation Techniques on Android Using Dynamic Analysis

Abstract

Android malware has become a growing issue for Android users, thanks to the popularity of the Operating System that has grown rapidly throughout the decade. Various approaches have been implemented to do malware detection in the Android OS, such as machine learning algorithms for heuristic approaches and frameworks to improve the detection and classification of malware. Unfortunately, most of the antivirus application that is available in the Android operating system is fully based on static methods, which cannot detect malware that used obfuscation techniques. This paper used the Mobile Security Framework by using the dynamic analysis method to detect malware with obfuscation techniques. The framework’s dynamic analysis method will be used to analyze multiple malware. The study aims to demonstrate dynamic analysis and measure the effectiveness of using dynamic analysis in detecting various malware. The method managed to detect 3 out of 7 malware applications that are present in the dataset. Several factors might contribute to the performance of the method, such as hardware limitations, obsolete application versions, and the ability of the application to behave differently with emulators.