Optimizing power control in centralized and distributed MIMO Networks: strategies and solutions

• Ano de defesa: 2024
• Autor(a) principal: Saraiva, Juno Vitorino
• Tipo de documento: Tese
• Tipo de acesso: Acesso aberto
• Idioma: eng
• Palavras-chave em Português: 5G Sistemas de Comunicação Móveis; Sistemas MIMO; Sistemas de comunicação móvel; 5G Mobile Communication Systems; MIMO Systems; Mobile Communication Systems; CNPQ::ENGENHARIAS::ENGENHARIA ELETRICA
• Link de acesso: http://repositorio.ufc.br/handle/riufc/78669

Definitely, the use of multiple input multiple output (MIMO) systems has provided significant benefits to mobile networks since previous generations. In this context, both centralized and distributed architectures of MIMO systems have been important to the evolution of cellular technologies, leading to significant improvements in key metrics such as spectral efficiency (SE) and energy efficiency (EE), fairness, and quality of service (QoS). However, two important challenges for current MIMO architectures are managing a massive number of connected users and efficiently integrating different types of users, such as ground user equipments (GUEs) and uncrewed aerial vehicles (UAVs), which have radically different mobility patterns and channel qualities. To address these challenges, radio resource management (RRM) solutions are essential. In this context, we propose various strategies to tackle different key objectives within mobile networks. More specifically, the first part of this thesis focuses on centralized MIMO networks and introduces RRM solutions leveraging fractional programming theory and game theory. Moreover, we consider scenarios that have been less explored in the literature, including scenarios with non-full buffer traffic models and time-correlated autoregressive channel models. In the second part of this thesis, we assume distributed MIMO networks. In this context, we initially adopt an approach under the category of potential games to propose a solution capable of attaining various network objectives, including SE, EE, and fairness. Subsequently, we explored scenarios involving the coexistence of GUEs and UAVs within the same network. Then, employing convex optimization and deep learning tools, we propose RRM solutions capable of accommodating diverse priority levels between GUEs and UAVs based on their individual requirements. For all RRM problems addressed in this thesis, decentralized solutions are proposed. Indeed, compared to centralized approaches, this becomes important as the number of users in the network grows indefinitely. Additionally, we aim to propose adaptable and flexible solutions that can easily adjust to changing objectives. This flexibility is crucial in dynamic and complex network environments.