Eigenvalue-Based Spectrum Sensing and Throughput in FD CR-IoT Networks Using Machine Learning

The Full-Duplex (FD) based Cognitive Radio-Internet of Things (CR-IoT) is an emerging technology for the Fifth Generation (5G) where simultaneously perform sensing and transmission in each time slot due to it mitigates the limitation of the Half-Duplex (HD) based CR-IoT networks. In this paper, we analysed the sensing and throughput based on the Eigenvalue-based Detection (EVD) technique using Machine Learning (ML) algorithms. Moreover, we analyzed the security issues based on the ML algorithms where separating the normal CR-IoT users from the Malicious Users (MUs). The simulation results show that greatly achieved better sensing gain, enhanced system throughput and enhanced security issue, compared to the conventional schemes.