Emerging and Unconventional Malware Detection and Mitigation Techniques

Abstract

Recent theoretical and practical studies have revealed that malware is one of the newlinemost harmful threats to the digital world. Malware mitigation techniques have evolved newlineover the years to ensure security. Earlier, several classical methods were used for de- newlinetecting malware embedded with various features like the signature, heuristic, and others. Traditional malware detection techniques were unable to defeat new generations of malware and their sophisticated obfuscation tactics. Today, developing a malware newlinedetection system is challenging, especially when dealing with new generation malware. newlineAdvanced evasion strategies have enabled the evolution of new generations of malware, newlinewhich had very significant effects. Objectives of this thesis are proposing a novel hybrid newlineanalysis technique to detect all kinds of malware and implementing the state of the art newlinesecurity solutions using password-less authentication. Four research contributions are provided for achieving the objectives. newlineThe second chapter offers a thorough analysis of the recently developed Deep Learn- newlineing based malware detection techniques in which current trending malware are studied newlineand detection techniques of Mobile malware (both Android and iOS), Windows mal- newlineware, IoT malware, Advanced Persistent Threats (APTs), and Ransomware are pre- newlinecisely reviewed. In the third chapter, the machine learning-based detection method newlineis designed by combining the features of application namely permission and activity newlinewhich are obtained during the installation of apps. newlineIn the fourth chapter, for the purpose of protecting physical assets of Internet of newlineMedical Things (IoMT) environment, Fast Identity Online Specifications (FIDO)2 spec- newlineifications and password-less authentication are incorporated. The fifth chapter proposes newlinea novel and robust decentralized financial transaction system, Painless Peer-to-peer Payments (PP2PP) that employs best cloud security practices, proper use of cryptography and trusted computing to mitigate common vulnerabilities. FIDO2 compatible Smart