Localização em ambiente interno usando a rede celular

• Ano de defesa: 2024
• Autor(a) principal: Conceição, Paulo Francisco da
• Orientador(a): Rocha, Flávio Geraldo Coelho
• Banca de defesa: Rocha, Flávio Geraldo Coelho, Vieira, Robson Domingos, Silva, Hugo Vinícius Leão e, Lemos, Rodrigo Pinto, Vieira, Flávio Henrique Teles
• Tipo de documento: Tese
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de Goiás
• Programa de Pós-Graduação: Programa de Pós-graduação em Engenharia Elétrica e da Computação (EMC)
• Departamento: Escola de Engenharia Elétrica, Mecânica e de Computação - EMC (RMG)
• País: Brasil
• Palavras-chave em Português: Localização; Rede celular; 5G; DCS-SOMP; ToA; AoD; AoA; Ambiente interno; SLAM
• Palavras-chave em Inglês: Localization; Cellular network; Indoor environment
• Área do conhecimento CNPq: ENGENHARIAS::ENGENHARIA ELETRICA
• Link de acesso: http://repositorio.bc.ufg.br/tede/handle/tede/13435

In this work, we propose an approach for Simultaneous Localization and Mapping (SLAM) of the Mobile Station (MS) and scatterers (SCs) in indoor environments using the cellular network. The approach, named IndoorLoc, employs a Single bounce scattering model and treats signals as rays, representing each radio frequency signal originating from multiple paths in an indoor environment where reflections occur at SCs. The estimation of the MS and SCs positions involves three main stages: (1) channel modeling, employing millimeter waves (mmWave) and a massive number of antennas (massive MIMO - mMIMO) arranged in a rectangular array; (2) parameter estimation, using an adaptive method based on Compressed Sensing (CS), the Distributed Compressed Sensing Simultaneous Orthogonal Matching Pursuit (DCS-SOMP); and (3) localization of the MS and SCs, applying geometric methods to detect Line of Sight (LoS) and Non-Line of Sight (NLoS) conditions and specific algorithms for each of these conditions. For IndoorLoc, two geometric methods are proposed: one for LoS conditions, which uses Time of Arrival (ToA) and Angle of Departure (AoD) parameters to determine the direction of propagation and the distance between the BS and the MS, and another for NLoS conditions, which uses ToA, AoD, and Angle of Arrival (AoA) to determine intersection points of the trajectories. This intersection point serves as the initial estimate of the MS localization and acts as an input for further refinement using a Gauss-Newton-based estimator, which minimizes the localization error using a nonlinear model derived from the ToA, AoD, and AoA parameters. The performance of the localization algorithms is evaluated through comparisons of the Root Mean Square Error (RMSE) values with existing methods in the literature. Additionally, simulations were conducted in an indoor environment configured according to the specifications of the 3rd Generation Partnership Project (3GPP). The results demonstrate that the accuracy of IndoorLoc meets the standards of the Federal Communications Commission (FCC) and 3GPP.