SILVA, L. J. V.; http://lattes.cnpq.br/9727531586375884; SILVA, Leandro José Ventura.
Resumo:
The world is currently in the era of Big Data. The predicted number of connected devices
to the Internet in the next decade is in the hundreds of billions. These devices range from
planes, cars, computers, cell phones to even small heart rate sensors. Processing these data volumes requires great computational powers. In addition, secure data processing is no longer a desirable requirement, increasing the computational cost. Thus, a software architecture model is proposed focusing on increasing the security of the processed data, in terms of reliability and integrity. This model is a specialization of the publish-subscribe standard with additional entities for access control, application and data integrity checking, and secure encryption key management. The proposed model was used for the development of two applications. The first application is in the area of smart grids, specifically in collecting household consumption information using smart meters. The second application is mobile data capture for use in targeted advertising. Both cases contain sensitive information that allowed to discover user behavior profiles. The main contributions of the model are data security guarantees, while scalable and compatibility across multiple clouds and non-cloud providers.