Performance Analysis of Non-Orthogonal Multiple Access (NOMA) in C-RAN, H-CRAN and F-RAN for 5G Systems

The world of telecommunication is witnessing a swift transformation towards fifth generation (5G) cellular networks. The future networks present requisite needs in ubiquitous throughput, low latency, and high reliability. They are also envisioned to provide diversified services such as enhanced Mobile BroadBand (eMBB) and ultra-reliable low-latency communication (URLLC) as well as improved quality of user experience. More interestingly, a novel mobile network architecture allowing centralized processing and cloud computing has been proposed as one of the best candidates for fifth generation. It is denoted as Cloud Radio Access Network (CRAN) and Heterogeneous Cloud Radio Access Network (H-CRAN). Furthermore, the 5G architecture will be fog-like, namely fog radio access networks (F-RAN) enabling a functional split of network functionalities between cloud and edge nodes with caching and fog computing capabilities. Meanwhile non-orthogonal multiple access (NOMA) has been proposed as an promising multiple access (MA) technology for future radio access networks (RANs) to meet the heterogeneous demands for high throughput, low latency and massive connectivity. One of the main challenges of NOMA is that how well it is to be compatible with other emerging techniques for meeting the requirements of 5G. However, comprehensive performance analysis on NOMA and practical resource allocation designs in co-existence with other emerging networks have not been fully studied and investigated in the literature. This thesis focuses on potential performance enhancement brought by NOMA for the C-RAN, H-CRAN and F-RAN and is expected to address some of the aforementioned key challenges of 5G. The research work of this thesis can be divided into three parts. In the first part of our research, we focus on investigating the performance analysis of NOMA in a C-RAN. The problem of jointly optimizing user association, muting and power-bandwidth allocation is formulated for NOMA-enabled C-RANs. To solve the mixed integer programming problem, the joint problem is decomposed into two subproblems as 1) user association and muting 2) power-bandwidth allocation optimization. To deal with the first subproblem, we propose a centralized and heuristic algorithm to provide the optimal and suboptimal solutions to the remote radio head (RRH) muting problem for given bandwidth and transmit power, respectively. The second subproblem is then reformulated and we propose an optimal solution to bandwidth and power allocation subject to users data rate constraints. Moreover, for given user association and muting states, the optimal power allocation is derived in a closed-form. Simulation results show that the proposed NOMA-enabled C-RAN outperforms orthogonal multiple access (OMA)-based C-RANs in terms of total achievable rate, interference mitigation and can achieve significant fairness improvement. Our second work investigates the performance of NOMA in H-CRAN, where coordination of macro base station (MBS) and remote radio heads (RRHs) for H-CRAN with NOMA is introduced to improve network performance. We formulate the problem of jointly optimizing user association, coordinated scheduling and power allocation for NOMA-enabled H-CRANs. To efficiently solve this problem, we decompose the joint optimization problem into two subproblems as 1) user association and scheduling 2) power allocation optimization. Firstly the users are divided based on different interference they suffer. This interference-aware NOMA approach account for the inter-tier interference. Proportional fairness (PF) scheduling for NOMA is utilized to schedule users with a two-loop optimization method to enhance throughput and fairness. Based on the user scheduling scheme, optimal power allocation optimization is performed by the hierarchical decomposition approach. It is then followed by algorithm for joint scheduling and power allocation. Simulation results show that the proposed NOMA-enabled H-CRAN outperforms OMA-based H-CRANs in terms of total achievable rate and can achieve significant fairness improvement. In the third part of our research, we propose a NOMA-enabled fog-cloud structure in a novel density-aware F-RAN to tackle different aspects such as throughput and latency requirements of high and low user-density regions, in order to meet the heterogeneous requirements of eMBB and URLLC traffic. A framework of the multi-objective problem is formulated to cater the high throughput and low-latency requirements in a high and low user-density mode respectively. In the first problem, we study the joint caching placement and association strategy aiming at minimizing the average delay. To deal with the first problem, we apply McCormick envelopes and Lagrange partial relaxation method to transform it into three convex sub-problems, which is then solved by proposed distributed algorithm. The second problem is to jointly optimize transmission mode selection, subchannel assignment and power allocation to maximize the sum data rate of all fog user equipments (F-UEs) while satisfying fronthaul capacity and fog-computing access point (F-AP) power constraints. Moreover, for given transmission mode selection and subchannel assignment, the optimal power allocation is derived in a closed-form. Simulation results are provided for the proposed NOMA-enabled F-RAN framework and reveal that the ultra-low latency and high throughput can be achieved by properly utilizing the available resources