Uma abordagem de Fog Computing para o subsistema de reconhecimento de contexto e adaptação do Middleware EXEHDA

Abstract

Recent surveys show that in the near future billions of smart devices will be interconnected via the Internet, thus attracting the attention of Industry and directing the research of the academic community, this synergy of investment has contributed to the materialization of the scenario known as IoT. From the perspective of IoT, computing provides information of all the "things" at all times, regardless of location, providing a highly distributed environment, heterogeneous, dynamic and strong interaction between man and machine. To this end, the IoT devices need to be aware of contextual data that interest you and where appropriate respond to, interoperating autonomously and with minimal human intervention possible in the aspects of management. For the processing of contextual data in IoT has been used cloud-based strategies, which has proven effective in the treatment of important aspects for the IoT, such as ease of access and availability. Howere, these strategies are vulnerable to systems that have limitations on their channels to the Internet, as well as for systems that require low latency in responses or present high disconnect chances. Given this motivation, the central objective of this thesis is the design of an architecture capable of providing the acquisition and processing of distributed contextual events. Therefore, the proposed architecture, called EXEHDA-FOG provides the middleware EXEHDA support to Fog Computing, using the distributed event processing at the edges as a cloud computing extension strategy. The results obtained with the case studies conducted have shown promising results, leading to the continuity of research efforts.