Detecção automática e alerta de acidentes de trânsito em redes veiculares reais

• Ano de defesa: 2015
• Autor(a) principal: Amarante, Thiago Carvalho
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Programa de Pós-Graduação em Ciência da Computação; UFLA; brasil; Departamento de Ciência da Computação
• Palavras-chave em Português: Redes veiculares; VANET; Vehicular networks; Ciência da computação
• Link de acesso: http://repositorio.ufla.br/jspui/handle/1/9769

Millions of people die annually around the world, victims of traffic accidents. In order to provide greater traffic safety, communication networks that can provide the exchange of information between vehicles was developed, the VANETs (Vehicular Ad Hoc Networks). This work presents as solution for traffic safety an application of automatic detection and warning for traffic accidents using VANETs. The transmission of accident warnings is done in an epidemic and opportunistic manner, spreading throughout all VANET devices found within the radius of interest to the accident information. We considered service quality requisites for the application of traffic safety and developed accident detection and warning dissemination algorithms in VANET. The application was evaluated in a real communication device. The results showed that it is possible to transmit warnings between vehicles and between vehicles and infrastructures. The warnings were transmitted with latency values below 100 milliseconds and with ranges superior to 150 meters in the evaluated sceneries.