Enhanced Permission Based Malware Detection in Mobile Devices Using Optimized Random Forest Classifier with PSO-GA

摘要

Smartphones and mobile devices are rapidly growing with their popularity as a part of global infrastructure powered communication system. As mobile devices become ubiquitous, used for a wide variety of application areas like personal communication, data storage, accessing online information, making payment, etc. The tremendous growth of smartphone usage makes it a target for malicious attackers to propagate malware attacks. Increased demand for mobile devices is due to the huge availability of applications that can be downloaded and installed easily on these devices. It is difficult for the general users to differentiate between the set of permissions which are potentially harmful and those which are not. This paper proposes to solve these issues by enhanced machine learning based malware detection framework using optimization algorithms. New classifier is developed by integrating GA and PSO with Random Forest algorithm. The Outcome from this paper is a new MSGP Malware Detection System consisting of MSGP-MDS Classifier. This reveals that classification of Android APK files using PSO plays a critical role in realizing higher accuracy with the minimum computation resource requirement.