Uma abordagem sensível a contexto para offloading de processamento

Abstract

The evolution of the processing and energy technologies of mobile devices has been quite accelerated in recent years. However, applications are following such evolution, requiring more processing power and energy capacity. Under these circumstances, approaches that aid in the processing of the applications become necessary. One of those is the approach of using the Fog Computing infrastructure, to speed up the execution of tasks on devices with low processing capacity and to avoid high latency of information exchange in the network. For this purpose, a process knew as offloading is used, where the processing of methods used by the applications is sent to other devices with the superior power of processing or energetic capacity. The delegation of these tasks could be done to a variety of remote execution environments (REE)s that have multiples characteristics that interfer with task completion speed. In this context, a decision method is needed to select where to perform offloading, once these REEs are heterogeneous and are dinamiccally connecting and desconnecting of the network. In this work, we propose a method to decide the best device to perform the offloading method, considering multiple criteria of these devices and their variance according to time, and present a functional infrastructure to enable the offloading of tasks by devices to others devices using the method proposed.