Hybrid Analysis Approach for Detecting Mobile Security Threats

Abstract

As technology continues to advance rapidly, smartphones are becoming increasingly powerful, drawing a large number of users with innovative features provided by mobile operating systems like Android. However, the security vulnerabilities of these systems make Android devices frequent targets for hackers and cyber criminals. As a result, research on effective and efficient mobile threat analysis has become a critical topic in the field of cyber security, employing methods such as static and dynamic analysis. This paper proposes a hybrid approach that combines static and dynamic analysis to detect security threats and attacks in mobile applications. The proposed method integrates data states and software execution along critical test paths. Initially, static analysis is used to identify potential attack paths based on Android APIs and existing attack patterns. This is followed by dynamic analysis, which executes the program along these paths within a focused scope to determine the likelihood of an attack by comparing detected paths with known attack patterns. In the runtime phase of dynamic analysis, the approach reports types of attack scenarios related to confidential data leakage, such as web browser cookies, while ensuring no actual critical or protected data on mobile devices is accessed.