Uma abordagem de processamento de consultas para plataformas de middleware distribuído e particionado no contexto de cidades inteligentes

• Ano de defesa: 2019
• Autor(a) principal: Lopes, Júlio Zinga Suzuki
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal da Paraíba; Brasil; Informática; Programa de Pós-Graduação em Informática; UFPB
• Palavras-chave em Português: Cidade inteligente; Linguagem de consulta de contexto; Sistema distribuído; Processamento de consulta; Middleware; Distributed systems; Query processing; Context query language; CNPQ::CIENCIAS EXATAS E DA TERRA::CIENCIA DA COMPUTACAO
• Link de acesso: https://repositorio.ufpb.br/jspui/handle/123456789/19444

Nowadays, more than half of the world's population lives in cities. It is expected to reach five billion people by 2030. This fast population growth in major centers will present biggest challenges, such as air pollution, urban mobility, health problems, energy and waste management. Solutions to these challenges require the integration of various Information and Communication Technologies (ICTs) in different domains. Cities that exploit these solutions, using ICTs to help solve their problems, have been called Smart Cities. However, the first solutions created had some limitations, as they were generally based on independent systems, with different technologies and without the concern with interoperability and scalability. In this context, middleware platforms have been used as infrastructure to assist the development of Smart City applications, systems and services. Among the main challenges of these platforms, data integration is very representative, as it allows users to provide consolidated information from different providers with transparency of location and representation format. In this sense, this paper proposes a query processing approach for distributed and partitioned middleware platforms, whose information is stored in different data providers and geographically distributed. A case study with significant volume of semi-public transit data was developed to validate and evaluate the proposed approach. In the performed case study, five distributed scenarios involving up to 10 platform servers, with 10,000 instances of context entities were used to evalute this proposal. In each scenario, the throughput in query processing was evaluated. Thus, compared to Orion Context Broker, a data provider widely used in European Smart Cities projects, the proposed approach was able to achieve a throughput exceeding 1.600%.