01 - Doutorado - Engenharia Elétrica

URI Permanente para esta coleção

http://repositorio.uel.br/handle/123456789/205

Navegar

Navegando 01 - Doutorado - Engenharia Elétrica por Autor "Abrão, Taufik"

Agora exibindo 1 - 3 de 3
  • Imagem de Miniatura
    Item
    Advanced Techniques for Channel Modeling, Estimation, and Resource Allocation Optimization in 5G/6G Wireless Communication Systems
    (2024-06-12) Guerra, David William Marques; Abrão, Taufik; Rego, Cássio Gonçalves; Panazio, Cristiano Magalhães; Filho, José Cândido Silveira Santos; Campos, Marcello Luiz Rodrigues
    Several emerging technologies have being proposed to meet the growing demand for highspeed, reliable, and high-quality communications for the next-generation (6G) mobile systems. The Extra-Large Multiple-Input Multiple-Output (XL-MIMO) system, a MIMO communication system with a very large number of antennas, is presented as a promising solution, as well as Reconfigurable Intelligent Surfaces (RIS), which are structures composed of reflecting elements capable of altering the phase and amplitude of the reflected signal, thus improving communication performance. Building on the potential of these techniques, this thesis explores the characteristics, challenges, and solutions associated with implementing these schemes in mobile communication systems. The first part proposes a double-scattering XL-MIMO channel modeling, considering specific concepts of this scenario and evaluating the impact of the spatial and temporal evolution of the dynamic environment. We observe that the birth and death processes of scatterers and clusters considerably impact system performance. In the second part, a channel estimator is proposed for RIS-assisted systems using signal compression techniques. More specifically, an efficient estimation of sparse channel gains using a modified redundant dictionary to reconstruct the BS-RIS/UE-RIS links in RIS-aided communication is proposed. This is achieved through a compressed sensing (CS) based method called Matching Pursuit with Phase Rotation (MP-PR), which uses a few active elements on the RIS panel. This procedure aims to optimize the efficiency and accuracy of channel estimation, representing one of the main challenges for effectively integrating RIS in beyond 5G (B5G) wireless networks. The third part of this work proposes an energy-efficient uplink power control for RIS-aided IoT systems. To address the battery limitations of IoT devices, the approach employs Riemannian manifolds and a convex power allocation solution to improve energy efficiency (EE) by reducing IoT transmit power in RIS-aided uplink Massive MIMO (M-MIMO) systems. Numerical results indicate significant improvements in the total power consumption of the devices compared to existing techniques. Additionally, an alternative approach using statistical channel state information (CSI) demonstrates comparable performance with reduced complexity
  • Imagem de Miniatura
    Item
    Eficiência de recursos sob equidade de taxa em sistemas NOMA, MIMO-NOMA e LSA-MIMO PDMA
    (2022-12-29) Jacob, Jaime Laelson; Abrão, Taufik; Vieira, Flávio Henrique Teles; Lima, Francisco Rafael Marques; Marinello Filho, José Carlos; Osorio, Diana Pamela Moya
    Analisa-se neste trabalho o compromisso entre a eficiência energética (EE - Energy Efficiency) e eficiência espectral (SE - Spectral Efficiency) de múltiplo acesso não ortogonal (NOMA - Non-Orthogonal Multiple Access), sistemas de múltiplas entradas e múltiplas saídas (MIMO - Multiple Input Multiple Output) combinados aos esquemas NOMA e MIMO-NOMA, bem como ao esquema de acesso múltiplo por divisão de padrões (PDMA - Pattern Division Multiple Acces) e ao MIMO massivo (mMIMO - Massive MIMO) em enlaces de descida (DL - Downlink) sob a restrição de equidade de taxa. Este compromisso entre EE e SE, denominado eficiência de recursos (RE - Resource Efficiency) é analisado para diferentes números de usuários por célula no NOMA. Obtêm-se o número de usuários por cluster na configuração MIMO-NOMA que resulta na máxima EE a partir da equidade de máxima taxa. Combinamos também o sistema mMIMO ao PDMA para oferecer suporte eficiente aos serviços de banda larga móvel avançada (eMBB - Enhanced Mobile Broadband), Comunicação Massiva do Tipo Máquina (mMTC - Massive Machine Type Communication). e Comunicações Ultra-Confiáveis e de Baixa Latência (uRLLC - Ultra-Reliable and Low Latency Communications) em sistemas da quinta geração (5G - Fifth Generation) e além da 5G (B5G - Beyond 5G); seleciona-se matrizes padrão que melhor atendem os requisitos heterogêneos desses serviços. A primeira parte desse trabalho trata da otimização da proporção de potência alocada para cada usuário nos sistemas NOMA. Na segunda parte, investigamos o compromisso EE-SE em sistemas MIMO-NOMA, considerando o DL, sob a restrição de mesma taxa entre os usuários. Essa restrição visa garantir a máxima equidade de taxa para todos os usuários ativos no sistema. Na terceira parte, o compromisso SE-EE foi formulado como um problema de otimização multi-objetivo (MOO - Multi-Objective Optimization) para o sistema MIMO-NOMA mantendo-se as restrições definidas anteriormente. O problema de otimização foi resolvido primordialmente adotando-se o método de escalarização ( ?- C – ?-Constraint) combinado a técnicas de programação não linear (NLP - Nonlinear Programming), particularmente a programação quadrática sequencial (SQP- Sequential Quadratic Programming) e o método de Dinkelback (DK). Finalmente, foram analisadas e comparadas várias matrizes PDMA disponíveis na literatura com características favoráveis ao atendimento dos requisitos impostos pelos diferentes serviços 5G, com uso da EE e outras métricas para a seleção da matriz padrão que melhor atende aos quesitos conflitantes em sistemas mMIMO PDMA. Resultados numéricos revelam que é possível identificar a melhor matriz padrão do sistema mMIMO PDMA que atenda requisitos conflitantes de cada um dos modos de uso 5G separadamente ou simultaneamente. No sistema NOMA, pode-se encontrar o ponto de equilíbrio entre EE e a soma de taxas para cada cenário do sistema, bem como o melhor ponto de operação de RE. Mostra-se ainda que o número total de usuários atinge a máxima EE em cada configuração cluster-usuários analisada. Resultados para a relação EE-SE no sistema MIMO-NOMA demonstraram a habilidade do método ?-C em encontrar diversidade de soluções na fronteira de Pareto
  • Imagem de Miniatura
    Item
    Physical layer optimization, multiple access, and energy efficiency in extra-large scale massive MIMO and wireless networks aided by reconfigurable intelligent surfaces
    (2024-03-01) Souza, João Henrique Inácio de; Abrão, Taufik; Klautau Júnior, Aldebaro Barreto da Rocha; Sadok, Djamel Fawzi Hadj; Rebelatto, João Luiz; Souza, Richard Demo; Marinello Filho, José Carlos
    The mobile networks beyond the fifth generation (B5G) must be designed to supply an increasing demand for connectivity, coming not only from established applications but also from emerging ones that surge with progressively ambitious communication requirements. To that end, physical layer technologies such as extra-large scale massive multiple-input multiple-output (XL-MIMO) and reconfigurable intelligent surfaces (RISs) have been considered for integrating the specifications of the next-generation networks aiming to boost the key performance indicators (KPIs). XL-MIMO encompasses communication systems where the base station is equipped with a physically large antenna array, with hundreds to thousands of half-wavelength-spaced antennas. Seen as an evolution of massive MIMO, in XL-MIMO systems, the array provides extra spatial degrees of freedom (DoFs) that can be explored to spatially multiplex many users with high data rates. Furthermore, an RIS is a thin sheet of composite material equipped with electronic circuits that can be programmed to change the characteristics of the incoming electromagnetic field. It can be used to coherently reflect radio signals, enhancing the wireless channel in arbitrary spots of the service area, relying on low-cost and energy-efficient hardware. To unlock the potential of XL-MIMO and RIS, it is necessary to solve open research questions concerning channel modeling, system design, and optimization strategies. Addressing the challenges faced in the development of these technologies, this thesis investigates the applicability of XL-MIMO and RIS to leverage the communication KPIs in the B5G networks. Based on the B5G usage scenarios, it focuses on four objectives: 1) providing high data rates and supporting high connection density for gigabit per second communication; 2) enabling energy-efficient wireless communication for the Internet of Things (IoT); 3) providing reliable and low-latency communication for mission-critical applications; and 4) investigating the potential of RISs to program aspects of the wireless channel. Regarding XL-MIMO, to support multi-user communication in crowded environments, a strategy for system optimization is proposed to efficiently explore all the DoFs provided by the high-aperture array. Furthermore, concerning RIS, the energy efficiency of RIS-aided IoT networks is analyzed, producing relevant insights for system design aiming to extend the devices’ battery lifetime and enhance the network coverage. Moreover, a multiple access scheme to multiplex hybrid traffic is proposed, using the RIS to support the coexistence through network resource sharing of mission-critical services along with broadband communication services. Finally, tackling the integration of the RIS controllability and exploring its capability of shaping the wireless channel, a method is proposed to control the channel temporal statistics by using an RIS with time-varying stochastic configurations. In short, this thesis presents methods, procedures, and algorithms to implement XL-MIMO systems and RISs in the B5G networks, accompanied by comprehensive evaluations of the system trade-offs in terms of relevant KPIs, including spectral efficiency, energy efficiency, outage probability, and latency.