When Machine Learning Meets Spectrum Sharing Security: Methodologies and Challenges

The exponential growth of Internet connected systems has generated numerous challenges, such as spectrum shortage issues, which require efficient sharing (SS) solutions. Complicated and dynamic SS can be exposed to different potential security privacy requiring protection mechanisms adaptive, reliable, scalable. Machine learning (ML) based methods have frequently been proposed address those issues. In this article, we provide a comprehensive survey the recent development ML methods, most critical corresponding defense mechanisms. particular, elaborate state-of-the-art methodologies for improving performance communication various vital aspects, including cognitive radio networks (CRNs), database assisted networks, LTE-U ambient backscatter other We also present issues from physical layer defending strategies on algorithms, Primary User Emulation (PUE) attacks, Spectrum Sensing Data Falsification (SSDF) jamming eavesdropping Finally, extensive discussions open challenges are given. This review is intended foundation facilitate future studies exploring emerging coping with increasingly complex their problems.