Cooperative localization improvement in vehicular ad hoc networks.

• Ano de defesa: 2020
• Autor(a) principal: Lobo, Felipe Leite
• Outros Autores: http://lattes.cnpq.br/1756041829894004
• Tipo de documento: Tese
• Tipo de acesso: Acesso aberto
• Idioma: eng
• Instituição de defesa: Universidade Federal do Amazonas; Instituto de Computação; Brasil; UFAM; Programa de Pós-graduação em Informática
• Palavras-chave em Português: Data Fusion; Localization Systems; Vehicular Ad-hoc Networks; Distance Information; Precise Localization System; CIÊNCIAS EXATAS E DA TERRA: CIÊNCIA DA COMPUTAÇÃO
• Link de acesso: https://tede.ufam.edu.br/handle/tede/7714

In Vehicular Ad Hoc Networks (VANets), a precise localization system is a crucial factor for several critical safety applications. Even though the Global Positioning System (GPS) can be used to provide the position estimation of vehicles, it still has an undesired error that can increase even more in some areas, such as tunnels and indoor parking lots, making it unreliable and unfeasible for most critical safety applications. In this work, we present a new position estimation technique by two algorithms, the CoVaLID (Cooperative Vehicle Localization Improvement using Distance Information), which improves GPS positions of nearby vehicles and minimize their errors using Extended Kalman Filter (EKF) to perform Data Fusion of both GPS and distance information, and the COLIDAP that uses Particle Filter (PF). Our solution also uses distance information to assess the position accuracy related to three different aspects: the number of vehicles, vehicle trajectory, and distance information error. For that purpose, we use a weighted average method to put more confidence in distance information given by neighbors closer to the target.We implement and evaluate the performance of CoVaLID using realworld data, as well as discuss the impact of different distance sensors in our proposed solution. Our results clearly show that our algorithms are capable of reducing the GPS error by 63%, and 53% when compared to the state-of-the-art VANet LOCation Improve (VLOCI) algorithm.