Uma abordagem de definição de políticas para offloading de dados no contexto de computação em névoa

Abstract

Currently, the most varied objects are connected to the Internet and, at the same time, generating massive amounts of data. Linked to this fact, the internet of things applications are increasingly complex and with more responsibilities. Storing, processing, managing, and analyzing this amount of data are challenging processes. The execution of these processes is commonly performed in external services through cloud computing, however, a paradigm called fog computing enables such execution directly at the edge of the network, serving as a support for the agile and efficient functioning of the internet of things. However, when fog computing does not have enough resources to perform these actions, the data is transferred to entities with higher computational capabilities, which is a practice known as offloading. In this regard, this research explores the use of policies that guide the process of data offloading in the context of fog computing. More specifically, the objective of this work is to define and organize strategies to guide the development of policies for data offloading in fog computing. Thus, the concrete results of the work were: the survey of policies for data offloading proposed by the literature; the development of a taxonomy that meets the main criteria used in the data offloading process; the development of a guide structure that recommends practices for policy making for data offloading and a demonstration of the instantiation of the approach, through proof of concept. Finally, the proof of concept found evidence that the approach provides more structured information for the design phase of data offloading policies and, consequently, contributes to their instance in real-world scenarios.