In the rapidly evolving landscape of wireless communication technologies, the deployment of the fifth generation(5G) mobile networks is underway, paving the path for the upcoming sixth generation (6G) that has recently captivated the interest of researchers in academia and industry. However, Mobile Network Operators (MNOs)have to face substantial challenges while deploying the 5G and the future 6G (collectively referred to as 5G and beyond), including cost, energy consumption, and user allocation concerns. The primary challenge is the high costs associated with network deployments, particularly in the fronthaul section, where the financial burden of establishing fronthaul networks becomes a pressing concern. Another challenge is the high energy consumption in 5G and beyond for environmental reasons and cost savings. Additionally, efficient user allocation is another crucial concern due to the diversity in user demands. Furthermore, ensuring continuous connectivity (higher availability)for the end user presents a significant challenge in maintaining uninterrupted services.
In this thesis, we propose novel optimization methods to address the aforementioned issues. First, we investigate the problem of cost-efficient optical fronthaul design for 5G and beyond. To do this properly, we propose an optimization framework based on Integer Linear Programming (ILP) for reducing the Total Cost of Ownership (TCO) and finding the optimal deployment of Time and Wavelength Division Multiplexing-Passive Optical Network (TWDM-PON)-based fronthaul. After that, we confirm the NP-hardness of the problem, and then we propose two heuristic methods based on K-means clustering algorithm and Genetic Algorithm (GA) for solving large-size problems.
Then, to choose the best optical fronthaul architecture, we provide a comparison in terms of cost and energy consumption for three optical architectures, namely: Point-to-Point fiber (P2P), PON, and hybrid PON-Free Space Optic (PON-FSO) when used for 5G and beyond fronthaul. We develop an end-to-end power consumption model for the studied architectures. Moreover, we propose an ILP formulation to reduce the TCO of the network when using each optical architecture. Then, to solve larger size problems, we introduce a heuristic approach named the Cost-Effective Optical Fronthaul Design Algorithm (CEOFDA).
Moving forward, the focus shifts to energy consumption and user allocation optimization in 5G millimeterWave (mmWave) networks. We investigate the problem of joint power and user allocation, representing it as an ILP. Then, we prove the NP-hardness of the problem and propose a heuristic method based on the GA. Then, we show the effectiveness of the proposed GA by comparing its performance with the ILP and two benchmark algorithms from the literature. Furthermore, given the diversity of user demands in 5G and beyond networks, we propose a Mixed Integer Linear Program (MILP) for effective resource allocation for three main user slices :Ultra-Reliable Low-Latency Communications (URLLC), massive Machine Type Communication (mMTC), and enhanced Mobile BroadBand (eMBB).
Considering the need for providing continuous connectivity for the end user simultaneously with reducing energy consumption, finally, in this thesis, we focus on the Dual Connectivity-User and Power Allocation (DC-UPA) problem in 5G mmWave networks. We represent the DC-UPA as an ILP and highlight its NP-hardness. Then, we propose two heuristic methods: the first is based on the Simulated Annealing (SA) algorithm, and the second is based on the Distance-Aware (DA) greedy algorithm. Simulations further confirm the efficiency of these methods against the ILP, considering two different user distributions (uniform and nonuniform).