Abstract

The fifth-generation (5G) communications and beyond are expected to serve a huge number of devices and services. However, due to the fixed spectrum allocation policies, the need for cognitive radio networks (CRNs) has increased accordingly. CRNs have been proposed as a promising approach to address the problem of under-utilization and scarcity of the spectrum. In CRNs, secondary users (SUs) access the licensed spectrum of the primary users (PUs) using underlay, overlay, or interweave paradigms. SUs can access the spectrum band simultaneously with the PUs in underlay access mode provided that the SUs’ transmission power does not cause interference to the PUs’ communication. In this case, SUs should keep monitoring the interference level that the PU receiver can tolerate and adjust the transmission power accordingly. However, varying the transmission power may lead to some threats to the privacy of the information transfer of CRNs. Therefore, securing data transmission in an underlay CRN is a challenge that should be addressed. Physical-layer security (PLS) has recently emerged as a reliable method to protect the confidentiality of the SUs’ transmission against attacks, especially for the underlay model with no need for sharing security keys. Indeed, PLS has the advantage of safeguarding the data transmission without the necessity of adding enormous additional resources, specifically when there are massively connected devices.
Apart from the energy consumed by the various functions carried out by SUs, enhancing security consumes additional energy. Therefore, energy harvesting (EH) is adopted in our work to achieve both; energy efficiency and spectral efficiency. EH is a significant breakthrough for green communication, allowing the network nodes to reap energy from multiple sources to lengthen battery life. The energy from various sources, such as solar, wind, vibration, and radio frequency (RF) signals, can be obtained through the process of EH. This accumulated energy can be stored to be used for various processes, such as improving the users’ privacy and prolonging the energy-constrained devices’ battery life.
In this thesis, for the purpose of realistic modelling of signal transmission, we explicitly assume scenarios involving moving vehicles or nodes in networks that are densely surrounded by obstacles. Hence, we begin our investigations by studying the link performance under the impact of cascaded κ−μ fading channels. Moreover, using the approach of PLS, we address the privacy of several three-node wiretap system models, in which there are two legitimate devices communicating under the threat of eavesdroppers. We begin by a three-node wiretap system model operating over cascaded κ − μ fading channels and under worst-case assumptions. Moreover, assuming cascaded κ − μ distributions for all the links, we investigate the impact of these cascade levels, as well as the impact of multiple antennas employed at the eavesdropper on security. Additionally, the PLS is examined for two distinct eavesdropping scenarios: colluding and non-colluding eavesdroppers. Throughout the thesis, PLS is mainly evaluated through the secrecy outage probability (SOP), the probability of non-zero secrecy capacity (Pnzcr ), and the intercept probability (Pint).
Considering an underlay CRN operating over cascaded Rayleigh fading channel, with the presence of an eavesdropper, we explore the PLS for SUs in the network. This study is then extended to investigate the PLS of SUs in an underlay single-input-multiple-output (SIMO) CRN over cascaded κ-μ general fading channels with the presence of a multi-antenna eavesdropper. The impact of the constraint over the transmission power of the SU transmitter due to the underlay access mode is investigated. In addition, the effects of multiple antennas and cascade levels over security are well-explored.
In the second part of our thesis, we propose an underlay CRN, in which an SU transmitter communicates with an SU destination over cascaded κ-μ channels. The confidentiality of the shared information between SUs is threatened by an eavesdropper. Our major objective is to achieve a secured network, while at the same time improving the energy and spectrum efficiencies with practical modeling for signals’ propagation. Hence, we presume that the SU destination harvests energy from the SU transmitter. The harvested energy is used to produce jamming signals to be transmitted to mislead the eavesdropper. In this scenario, a comparison is made between an energy-harvesting eavesdropper and a non-energy harvesting one. Additionally, we present another scenario in which cooperative jamming is utilized as one of the means to boost security. In this system model, the users are assumed to communicate over cascaded Rayleigh channels. Moreover, two scenarios for the tapping capabilities of the eavesdroppers are presented; colluding and non-colluding eavesdroppers. This study is then extended for the case of non-colluding eavesdroppers, operating over cascaded κ-μ channels.
Finally, we investigate the reliability of the SUs and PUs while accessing the licensed bands using the overlay mode, while enhancing the energy efficiency via EH techniques. Hence, we
assume that multiple SUs are randomly distributed, in which one of the SUs is selected to harvest energy from the PUs’ messages. Then, utilizing the gathered energy, this SU combines its own
messages with the amplified PUs messages and forwards them to the destinations. Furthermore, we develop two optimization problems with the potential of maximizing the secondary users’ rate and the sum rate of both networks.