Design and Implementation of a Malware Detection Tool Using Network Traffic Analysis in Android-based Devices

Abstract

Smartphone use and dependability have increased substantially in recent years, as have malicious attempts to compromise mobile devices with various malware. Therefore, smartphones must have an active malware detection program to protect user privacy. We propose Android Malware Buster (AMB), a malware detection application for Android devices. AMB utilizes a machine learning classifier to identify ongoing malicious behavior through its network traffic analysis. The machine learning model was trained on a diverse set of Adware, Scareware, and Ransomware apps. The accuracy of the AMB classifier has reached 93 %. Furthermore, AMB correctly classified the vast majority of applications during real-time testing.