Um mecanismo de offloading para redes de sensores veiculares

• Ano de defesa: 2019
• Autor(a) principal: Douglas Leite Leal Moura
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de Minas Gerais; Brasil; ICX - DEPARTAMENTO DE CIÊNCIA DA COMPUTAÇÃO; Programa de Pós-Graduação em Ciência da Computação; UFMG
• Palavras-chave em Português: Redes veiculares; VANETs; Offloading; Computação – Teses; Redes veiculares -Teses; VANETs - Teses; Offloading - Teses
• Link de acesso: http://hdl.handle.net/1843/40046

The vehicular ad-hoc network is an important data source for a remote sensing paradigm known as vehicular sensor networks. Vehicular sensor networks have been widely studied and proved to be very usefull for urban environmental monitoring, providing information services, efficient traffic management and long-term urban planning. The acquisition of a massive amount of sensing data can generate a significant increase in cellular traffic and reduce network capacity, therefore strategies are required in order to provide bandwidth savings and scalable network management. This work presents an offloading approach for sensing data in a vehicular sensor networks. The data that would originally be transmitted in the cellular network will be migrated to a complementary and low-cost network, based on D2D communication, to be compressed by some data reduction technique and transmitted to the base station. We model the problem as an instance of d-hop dominating set and propose a centrality based algorithm to determine a subset of vehicles to perform aggregation and upload of the data. The experiments used mobility traces from real scenario and results show a decrease of up to 155.40 kB/s in the upload cost during peak hours.