Using Taint Analysis for Threat Risk of Cloud Applications

—Most existing approaches to developing cloud applications using threat analysis involve program vulnerability analyses for identifying the security holes associated with malware attacks. New malware attacks can bypass firewall-based detection by bypassing stack protection and by using Hypertext Transfer Protocol logging, kernel hacks, and library hack techniques, and to the cloud applications. In performing threat analysis for unspecified malware attacks, software engineers can use a taint analysis technique for tracking information flows between attack sources (malware) and detect vulnerabilities of targeted network applications. This paper proposes a threat risk analysis model incorporating an improved attack tree analysis scheme for solving the mobile security problem; in the model, Android programs perform taint checking to analyse the risks posed by suspicious applications. In probabilistic risk analysis, defence evaluation metrics are used for each attack path for assisting a defender simulate the attack results against malware attacks and estimate the impact losses. Finally, a case of threat analysis of a typical cyber security attack is presented to demonstrate the proposed approach.