CYBER THREAT INTELLIGENCE FOR ANDROID ECOSYSTEMS: A MATHEMATICAL FRAMEWORK FOR ADVANCED DETECTION OF HIDDEN MALICIOUS ATTACKS IN MOBILE APPLICATIONS

Abstract

Android mobile application has emerged among the best targets of the advanced cyber threats which aim at evading their use through the sophisticated evasion techniques so as not to be detected within the walls of the applications that seem to be legal. This paper proposes an elaborate mathematical model of cyber threat intelligence in detecting the obfuscated malicious attacks within the Android environment. We discuss some of the cyber-attack vectors including code obfuscation, dynamic loading and permission abuse where bad code can bypass traditional cyber security controls. To tackle these cyber crimes, two mathematically based algorithms are suggested which can be applied in combating these crimes, namely Permission Anomaly Detection Algorithm (PADA) and Mathematical Behavioral Analysis Algorithm (MBAA). In a more empirical sense, using our integrated mechanism in detecting cyber threats in the environment of 1,000 Android applications, our detection percentage on different cyber threat scenario is a 94.2 with a false positive of only 3.1 percent. Its outcomes are significant in regards to mathematical intelligence of cyber threats and how to create more resilient security models of mobile application environments which happen to be faced with the problem of cyber-attacks on Android platforms.