| Resumo: | Mobile services incrementally demand for further processing and storage. However, mobile devices are known for their constrains in terms of processing, storage, and energy. Early proposals have addressed these aspects; by having mobile devices access remote clouds. But these proposals suffer from long latencies and backhaul bandwidth limitations in retrieving data. To mitigate these issues, edge clouds have been proposed. Using this paradigm, intermediate nodes are placed between the mobile devices and the remote cloud. These intermediate nodes should fulfill the end users’ resource requests, namely data and processing capability, and reduce the energy consumption on the mobile devices’ batteries. But then again, mobile traffic demand is increasing exponentially and there is a greater than ever evolution of mobile device’s available resources. This urges the use of mobile nodes’ extra capabilities for fulfilling the requisites imposed by new mobile applications. In this new scenario, the mobile devices should become both consumers and providers of the emerging services. The current work researches on this possibility by designing, implementing and testing a novel nomadic fog storage system that uses fog and mobile nodes to support the upcoming applications. In addition, a novel resource allocation algorithm has been developed that considers the available energy on mobile devices and the network topology. It also includes a replica management module based on data popularity. The comprehensive evaluation of the fog proposal has evidenced that it is responsive, offloads traffic from the backhaul links, and enables a fair energy depletion among mobiles nodes by storing content in neighbor nodes with higher battery autonomy.
|
|---|