Energy Harvesting-enabled Cognitive Radio-Internet of Things Using Machine Learning

Abstract

The cognitive radio (CR) is a modern technology in cognitive radio-Internet of things (CR-IoT) networks. In contrast, each CR-IoT user is unable to achieve both a better sensing gain, and an enhanced system sum rate in conventional energy detection technique based CR-IoT networks with the present energy harvesting (EH) and security threats due to under-utilized the reporting framework. For this reason, we proposed EH-enabled CR-IoT networks using machine learning (ML) algorithms in which each normal CR-IoT user is assisted by finite capacity battery systems and energy harvested. In this paper, Firstly, the proposed hybrid detection technique based on EH-enabled CR-IoT networks using ML algorithms is separating the trusted (normal) and untrusted (malicious users) CR-IoT users where all untrusted CR-IoT users are not participating in spectrum sensing due to they degraded the performance like sensing gain and system sum rate; Secondly, the proposed scheme is utilized the reporting framework where only trusted CR-IoT users are obtained longer sensing time slot which enhanced the sensing gain, the EH, and the sum rate; and Finally, the simulation results show that the proposed hybrid scheme outperformed the conventional schemes in terms of security, sensing gain, EH, and system sum rate.