In underlay cognitive radio networks, secondary users can share the spectrum with primary users as long as the interference caused by the secondary users to primary users is below a certain predetermined threshold. It is reasonable to assume that there is always a large pool of secondary users trying to access the channel, which can be occupied by only one secondary user at a given time. As a result, a multi-user scheduling problem arises among the secondary users. In this thesis, by manipulating basic schemes based on selective multi-user diversity, normalized thresholding, transmission power control, and opportunistic round robin, we propose and analyze eight scheduling schemes of secondary users in an underlay cognitive radio set-up. The system performance of these schemes is quantified by using various performance metrics such as the average system capacity, normalized average feedback load, scheduling outage probability, and system fairness of access. In our proposed schemes, the best user out of all the secondary users in the system is picked to transmit at each given time slot in order to maximize the average system capacity. Two thresholds are used in the two rounds of the selection process to determine the best user. The first threshold is raised by the power constraint from the primary user. The second threshold, which can be adjusted by us, is introduced to reduce the feedback load. The overall system performance is therefore dependent on the choice of these two thresholds and the number of users in the system given the channel conditions for all the users. In this thesis, by deriving analytical formulas and presenting numerical examples, we try to provide insights of the relationship between the performance metrics and the involved parameters including two selection thresholds and the number of active users in the system, in an effort to maximize the average system capacity as well as satisfy the requirements of scheduling outage probability and feedback load.

In the first part of the dissertation, we provide an extensive overview of sub-6 GHz wireless access technology known as massive multiple-input multiple-output (MIMO) systems, highlighting its benefits, deployment challenges, and the key enabling technologies envisaged for 5G networks. We investigate the fundamental issues that degrade the performance of massive MIMO systems such as pilot contamination, precoding, user scheduling, and signal detection. In the second part, we optimize the performance of the massive MIMO system by proposing several algorithms, system designs, and hardware architectures. To mitigate the effect of pilot contamination, we propose a pilot reuse factor scheme based on the user environment and the number of active users. The results through simulations show that the proposed scheme ensures the system always operates at maximal spectral efficiency and achieves higher throughput. To address the user scheduling problem, we propose two user scheduling algorithms bases upon the measured channel gain. The simulation results show that our proposed user scheduling algorithms achieve better error performance, improve sum capacity and throughput, and guarantee fairness among the users. To address the uplink signal detection challenge in the massive MIMO systems, we propose four algorithms and their system designs. We show through simulations that the proposed algorithms are computationally efficient and can achieve near-optimal bit error rate performance. Additionally, we propose hardware architectures for all the proposed algorithms to identify the required physical components and their interrelationships.

[pt] As redes multi-input multi-output massivas livres de células (CFmMIMO) são uma evolução promissora nas comunicações sem fio, oferecendo melhorias notáveis na experiência do usuário e no desempenho da rede ao eliminar as fronteiras tradicionais das células. Essas redes empregam um grande número de pontos de acesso (APs) distribuídos para servir um número menor de equipamentos de usuário (UEs), formando uma arquitetura de rede sem fio única que garante uma cobertura sólida e entrega de serviço. Uma preocupação primária no contexto das redes CF-mMIMO é a alocação eficiente de recursos, particularmente o agendamento de usuários e a alocação de potência. O objetivo desta dissertação é investigar essas tarefas na transmissão downlink de uma rede CF-mMIMO, considerando tanto a informação perfeita quanto a imperfeita do estado do canal (CSI). No agendamento de usuários, o objetivo é selecionar um subconjunto de UEs para serem atendidos em um determinado momento. Esse processo é inerentemente complexo, pois deve levar em consideração diversos fatores, como a prioridade dos UEs, as condições variáveis do canal e as localizações físicas dos UEs. O agendamento eficaz de usuários é fundamental para otimizar a utilização dos recursos da rede, melhorar a satisfação dos UEs e gerenciar o tráfego da rede de forma eficiente. A alocação de potência também desempenha um papel crucial na distribuição da potência de transmissão entre os APs e os UEs selecionados. Nos sistemas CF-mMIMO, uma estratégia eficaz de alocação de potência pode ajudar a mitigar a interferência entre usuários e otimizar a eficiência energética, enquanto impõe as restrições de potência total disponível e de hardware. Esta dissertação também considera o contexto das redes livres de células agrupadas (CLCF), como o agrupamento de rede não sobreposto e também o livres de células centrado no usuário (UCCF). Essas redes apresentam seus próprios desafios e oportunidades únicos em termos de agendamento de usuários e alocação de potência. Uma exploração e comparação aprofundada de diferentes técnicas dentro desses contextos poderia oferecer insights valiosos para o desenvolvimento de estratégias de alocação de recursos mais eficientes. A dissertação visa fornecer um estudo abrangente da alocação de recursos, focando no agendamento de usuários e na alocação de potência na transmissão downlink de redes CF-mMIMO, levando em consideração tanto a CSI perfeita quanto a imperfeita e explorando as implicações dessas técncas nos contextos das redes CLCF e UCCF, além de indicar alguns tópicos para investigação futura. / [en] Cell-Free Massive Multiple-Input Multiple-Output (CF-mMIMO) networks are a promising evolution in wireless communications, offering notable improvements in user experience and network performance by eliminating traditional cell boundaries. These networks employ a large number of distributed access points (APs) to serve a smaller number of user equipments (UEs), forming a unique wireless network architecture that ensures solid coverage and service delivery. A primary concern in the context of CF-mMIMO networks is the efficient allocation of resources, particularly user scheduling and power allocation. The aim of this thesis is to investigate these tasks in the downlink of a CF-mMIMO network, considering both perfect and imperfect channel state information (CSI). In user scheduling, the objective is to select a subset of UEs to be served at any given time. This process is inherently complex as it must cater for numerous factors such as the priority of UEs, varying channel conditions, and physical UE locations. Effective user scheduling is instrumental in optimizing the utilization of network resources, enhancing UE satisfaction, and managing network traffic efficiently. Power allocation also plays a key role in the distribution of transmission power among APs and the selected UEs. In CF-mMIMO systems, an effective power allocation strategy can help mitigate inter-user interference and optimize energy efficiency, while enforcing the total available power and hardware constraints. This thesis further considers the context of clustered cell-free (CLCF) networks, as non-overlapping network clustering and also usec-centric cell-free (UCCF). These networks present their own unique challenges and opportunities in terms of user scheduling and power allocation. An in-depth exploration and comparison of different techniques within this settings could offer valuable insights into the development of more efficient resource allocation strategies. The thesis aims to provide a comprehensive study of resource allocation, focusing on user scheduling and power allocation in the downlink of CF-mMIMO networks, taking into account both perfect and imperfect CSI and exploring the implications of these techniques in CLCF and UCCF network contexts, and indicate some topics for further investigation in the future works.