Abstract

With the proliferation of numerous burgeoning services such as ultra-reliable low-latency communication (URLLC), massive machine type communications (mMTC), enhanced mobile broadband (eMBB), among others, wireless communication systems are expected to face daunting challenges. In order to satisfy these ever-increasing traffic demands, diverse quality-of-services (QoS) requirements, and the massive connectivity accompanied by these new applications, various innovative and promising technologies, and architectures need to be developed. Novel multiple-access techniques are currently being explored in both academia and industry in order to accommodate such unprecedented requirements. Non-orthogonal multiple access (NOMA) has been deemed as one of the vital enabling multiple access techniques for the upcoming six-generation (6G) networks. This is due to its ability to enhance network spectral efficiency (NSE) and support a massive number of connected devices. Owing to its potential benefits, NOMA is recognized as a prominent member of next-generation multiple access (NGMA).

Several emerging techniques such as full-duplex (FD) communication, device-to-device (D2D) communications, reconfigurable intelligent surface (RIS), coordinated multipoint (CoMP), cloud radio access networks, are being gradually developed to address fundamental problems in future wireless networks. In this thesis, and with the goal of converging toward NGMA, we investigate the synergistic integration between NOMA and other evolving physical layer technologies. Specifically, we analyze this integration aiming at improving the performance of cell-edge users (CEUs), mitigating the detrimental effect of inter-cell interference (ICI), designing energy-efficient multiple access toward ``green’’ wireless networks, guarantying reliable communication between NOMA UEs and base stations (BSs)/remote radio heads (RRHs), and maintaining the required QoS in terms of the minimum achievable data rate, especially at CEUs.

Regarding the ICI mitigation in multi-cell NOMA networks and tackling the connectivity issue in traditional CoMP-based OMA networks, we first investigate the integration between location-aware CoMP transmission and NOMA in downlink heterogeneous C-RAN. In doing so, we design a novel analytical framework using tools from stochastic geometry to analyze the system performance in terms of the average achievable data rate per NOMA UE. Our results reveal that CoMP NOMA can provide a significant gain in terms of network spectral efficiency compared to the traditional CoMP OMA scheme. In addition, with the goal of further improving the performance of CEUs and user fairness, cooperative transmission with the aid of D2D communication and FD or half-duplex (HD) transmission, has been introduced to NOMA, which is commonly known as cooperative NOMA (C-NOMA). As a result, we extend our study to also investigate the potential gains of investigating CoMP and C-NOMA. In such a framework, we exploit the cooperation between the RRHs/BSs and the successive decoding strategy at NOMA UEs that are near the RRHs/BSs. Specifically, we investigate both performance analysis and resource management optimization (power control and user pairing). Our results show that the transmit power at the BS, the transmit power at the relay user, and the self-interference (SI) value at the relay user determine which multiple access technique, CoMP NOMA, CoMP HD C-NOMA, and CoMP FD C-NOMA, should be adopted at the BSs.

Now, to assist in designing energy-efficient multiple access techniques and guarantying reliable communication for NOMA UEs, this thesis explores the interplay between FD/HD C-NOMA and RIS. We show that the proposed model has the best performance in terms of network power consumption compared to other multiple access techniques in the literature, which leads to ``green'' future wireless networks. Moreover, our results show that the network power consumption can be significantly reduced by increasing the number of RIS elements. A more significant finding is that the location of the RIS depends on the adopted multiple access techniques. For example, it is not recommended to deploy the RIS besides the BS if the adopted multiple access is HD C-NOMA. Another insight that has been unveiled is the FD C-NOMA with the assistance of RIS has more resistance to the residual SI effect, due to the FD transmission, and can tolerate high SI values compared to the same scheme without RIS.

Although much work has been conducted to improve the network spectral efficiency of multi-cell NOMA cellular networks, the required QoS by the upcoming 6G applications, in terms of the minimum achievable rate, may not be guaranteed at CEUs. This is due to their distant locations from their serving BSs, and thus, they experience severe path-loss attenuation and high ICI. This thesis addresses this research gap by studying the synergistic integration between RIS, NOMA, and CoMP in a multi-user multi-cell scenario. Unlike the developed high-complexity optimal solutions or the low-complexity sub-optimal solutions in the literature for the power allocation problem, we derive a low-complexity optimal solution in a such challenging scenario. We also consider the interdependency between the user clustering policies in different coordinated cells, which has been ignored in the literature. Finally, we prove that this integration between RIS, NOMA, and CoMP can attain a high achievable rate for CEUs, ameliorate spectral efficiency compared to existing literature, and can form a novel paradigm for NGMA.