Android botnet categorization and family detection based on behavioural and signature data

Predicting application performing malicious activity based on its behavioural analysis is extremely difficult compare to signature based approach. But considering the rapid development and slight changes in code allowing avoiding of signature-based malware analysis has made behaviour-based analysis more and more important in recent years. In last decade there is unimagined and trilling growth in the mobile market, which is unquestionably dominated by Android OS. Android has very fast growing application markets that have been targeted by underground malware distribution networks. There are larger numbers of new application stores across the globe apart from leaders of App market like Google Play Store, Amazon etc. It is very important to test the possible ways of behaviour based malware analysis in Android. Research focuses on creating a working prototype of a system that takes different behavioural parameters of Android application as input and perform analysis using artificial intelligence approach. During the implementation signature based detection methods were also included in the implemented prototype.