The cognitive radio (CR) concept is expected to be adopted along with many
technologies to meet the requirements of the next generation of wireless and mobile
systems, the 5G. Consequently, it is important to determine the performance of the
CR systems with respect to these requirements. In this thesis, after briefly describing
the 5G requirements, we present three main directions in which we aim to enhance
the CR performance.
The first direction is the reliability. We study the achievable rate of a multiple-input multiple-output (MIMO) relay-assisted CR under two scenarios; an unmanned
aerial vehicle (UAV) one-way relaying (OWR) and a fixed two-way relaying (TWR).
We propose special linear precoding schemes that enable the secondary user (SU) to
take advantage of the primary-free channel eigenmodes. We study the SU rate sensitivity to the relay power, the relay gain, the UAV altitude, the number of antennas
and the line of sight availability.
The second direction is the scalability. We first study a multiple access channel
(MAC) with multiple SUs scenario. We propose a particular linear precoding and SUs
selection scheme maximizing their sum-rate. We show that the proposed scheme provides a significant sum-rate improvement as the number of SUs increases. Secondly, we expand our scalability study to cognitive cellular networks. We propose a low-complexity algorithm for base station activation/deactivation and dynamic spectrum
management maximizing the profits of primary and secondary networks subject to green constraints. We show that our proposed algorithms achieve performance close to those obtained with the exhaustive search method.
The third direction is the energy efficiency (EE). We present a novel power allocation scheme based on maximizing the EE of both single-input and single-output
(SISO) and MIMO systems. We solve a non-convex problem and derive explicit expressions of the corresponding optimal power. When the instantaneous channel is not available, we present a simple sub-optimal power that achieves a near-optimal EE.
The simulations show that the sub-optimal solution is very close to the optimal one.
In the MIMO case, we show that adopting more antennas is more energy efficient.
|Date of Award||Nov 2017|
|Original language||English (US)|
- Computer, Electrical and Mathematical Science and Engineering
|Supervisor||Mohamed-Slim Alouini (Supervisor)|
- Achievable Rate
- Cognitive Radio
- Energy Efficiency
- Green cellular networks
- MIMO communications
- Relaying communications