Drone-Assisted Wireless Communications

In order to address the increased demand for any-time/any-where wireless connectivity, both academic and industrial researchers are actively engaged in the design of the fifth generation (5G) wireless communication networks. In contrast to the traditional bottom-up or horizontal design approaches, 5G wireless networks are being co-created with various stakeholders to address connectivity requirements across various verticals (i.e., employing a top-to-bottom approach). From a communication networks perspective, this requires obliviousness under various failures. In the context of cellular networks, base station (BS) failures can be caused either due to a natural or synthetic phenomenon. Natural phenomena such as earthquake or flooding can result in either destruction of communication hardware or disruption of energy supply to BSs. In such cases, there is a dire need for a mechanism through which capacity short-fall can be met in a rapid manner. Drone empowered small cellular networks, or so-called \quotes{flying cellular networks}, present an attractive solution as they can be swiftly deployed for provisioning public safety (PS) networks. While drone empowered self-organising networks (SONs) and drone small cell networks (DSCNs) have received some attention in the recent past, the design space of such networks has not been extensively traversed. So, the purpose of this thesis is to study the optimal deployment of drone empowered networks in different scenarios and for different applications (i.e., in cellular post-disaster scenarios and briefly in assisting backscatter internet of things (IoT)). To this end, we borrow the well-known tools from stochastic geometry to study the performance of multiple network deployments, as stochastic geometry provides a very powerful theoretical framework that accommodates network scalability and different spatial distributions. We will then investigate the design space of flying wireless networks and we will also explore the co-existence properties of an overlaid DSCN with the operational part of the existing networks. We define and study the design parameters such as optimal altitude and number of drone BSs, etc., as a function of destroyed BSs, propagation conditions, etc. Next, due to capacity and back-hauling limitations on drone small cells (DSCs), we assume that each coverage hole requires a multitude of DSCs to meet the shortfall coverage at a desired quality-of-service (QoS). Hence, we consider the clustered deployment of DSCs around the site of the destroyed BS. Accordingly, joint consideration of partially operating BSs and deployed DSCs yields a unique topology for such PS networks. Hence, we propose a clustering mechanism that extends the traditional Mat\'{e}rn and Thomas cluster processes to a more general case where cluster size is dependent upon the size of the coverage hole. As a result, it is demonstrated that by intelligently selecting operational network parameters such as drone altitude, density, number, transmit power and the spatial distribution of the deployment, ground user coverage can be significantly enhanced. As another contribution of this thesis, we also present a detailed analysis of the coverage and spectral efficiency of a downlink cellular network. Rather than relying on the first-order statistics of received signal-to-interference-ratio (SIR) such as coverage probability, we focus on characterizing its meta-distribution. As a result, our new design framework reveals that the traditional results which advocate lowering of BS heights or even optimal selection of BS height do not yield consistent service experience across users. Finally, for drone-assisted IoT sensor networks, we develop a comprehensive framework to characterize the performance of a drone-assisted backscatter communication-based IoT sensor network. A statistical framework is developed to quantify the coverage probability that explicitly accommodates a dyadic backscatter channel which experiences deeper fades than that of the one-way Rayleigh channel. We practically implement the proposed system using software defined radio (SDR) and a custom-designed sensor node (SN) tag. The measurements of parameters such as noise figure, tag reflection coefficient etc., are used to parametrize the developed framework.