Data and Task Offloading in Collaborative Mobile Fog-Based Networks

The rapid growth of mobile devices in recent years has led to the rapid progress of mobile computing. However, this has exposed certain limitations that have, first, been addressed by mobile cloud computing. Once the Internet-of-Things devices have started being put online, a new step in the evolution of mobile networks was taken through the addition of edge and fog computing, where small nodes at the edge of the network take up some of the load on the cloud backend. Nevertheless, even this model has shown some limitations, which is why in this paper, we address the problem of off-loading data and computations from a mobile device to the cloud to fog nodes, or to other mobile devices in the vicinity. The novelty of our proposal is the addition of a layer composed exclusively of mobile devices that collaborate in an opportunistic fashion, as a first resort when needing some computations to be off-loaded. Through a thorough analysis using the MobEmu mobile network simulator, we show that our solution is able to reduce total computation time by as much as 19%, decrease the cloud usage with up to 40%, and reduce battery consumption with more than 6%.