A Survey on Various Threats and Current State of Security in Android Platform

Abstract

The advent of the Android system has brought smartphone technology to the doorsteps of the masses. The latest technologies have made it affordable for every section of the society. However, the emergence of the Android platform has also escalated the growth of cybercrime through the mobile platform. Its open source operating system has made it a center of attraction for the attackers. This article provides a comprehensive study of the state of the Android Security domain. This article classifies the attacks on the Android system in four categories (i) hardware-based attacks, (ii) kernel-based attacks, (iii) hardware abstraction layer-based attacks, and (iv) application-based attacks. The study deals with various threats and security measures relating to these categories and presents an in-depth analysis of the underlying problems in the Android security domain. The article also stresses the role of Android application developers in realizing a more secure Android environment. This article attempts to provide a comparative analysis of various malware detection techniques concerning their methods and limitations. The study can help researchers gain knowledge of the Android security domain from various aspects and build a more comprehensive, robust, and efficient solution to the threats that Android is facing.

A Review of Android Malware Detection Approaches Based on Machine Learning

“…The analysis methods used to obtain these three types of features are called static analysis, dynamic analysis, and hybrid analysis, respectively. The static analysis method analyzes the application and related objects without executing the application [35]. Most static methods use techniques that parse program source code to traverse program paths to check some properties [36].…”

“…The dynamic analysis method analyzes features while the application is running (on a real device or virtual environment) [35]. Dynamic analysis is mainly used to identify the behavior characteristics of an Android application, and techniques such as function call monitoring, information flow tracking, and instruction tracing can be applied [17], [24].…”

“…In this case, time bombs, logic bombs, and login interfaces based on passwords, for example, will bring difficulties to our automated analysis. Although a lot of research work has put forward methods to address these deficiencies, it is still difficult to achieve a perfect unity between the analysis effectiveness and efficiency [4], [35], [37].…”

“…As shown in Table 1, this review differs from several previous works in this discipline. It is not a general study of Android malware detection [15], [23] or Android security [35], [37] but instead systematically focuses on key perspectives on machine learning approaches used in Android malware detection. In some papers, although machine learning methods are included, the information is scattered throughout the article to support different objectives.…”

Detection, Traceability, and Propagation of Mobile Malware Threats

“…Due to the open nature of Android devices, mobile device manufacturers are rapidly producing various Android versions worldwide. Android mobile malware is enriched in the cumbersome app store download and installation authorization scenarios [11].…”