A High-Fidelity Detection Strategy for Android Malware based on Co-Existing Features

The smartphone market is expanding tremendously, according to the ICD (International Code Designator) report. Android is a popular mobile operating system. Concerns regarding device security and personal privacy have significantly increased due to the extensive variety of features smartphones offer and the growing number of activities they enable, such as online social interaction, digital financial services, and internet gaming. Because Android is an open-source platform, malware developers can easily launch attacks and create dangerous Android malware applications. To evade detection by anti-malware scanners, malware is continuously updated with evolving technology. To gain access to a user's device, Android malware programs often employ three breakthrough tactics: downloading, updating, and repackaging. This study presents a unique machine learning model that integrates static attribute patterns to identify Android malware. It is based on the assumption that, unlike benign apps, Android malware seeks an unusual combination of coexisting permissions and APIs. Several traditional learning methods were employed to assess the proposed model. The findings demonstrate that co-existence of characteristics is a useful technique for identifying Android malware.