A Platform for Distributed Fog Computing

Abstract

The fog computing paradigm has introduced the concept of processing data near the data source. Unlike the cloud, fog computing includes devices with highly varied resources such as heterogeneous computing power, battery, bandwidth, delay, and mobility. However, existing distributed computing frameworks and storage systems have mainly focused on the cloud environment where resources are highly consolidated and stable. This thesis presents Fogos, a platform for a distributed fog computing framework and fog storage system. In addition, a decentralized incentive system is proposed to encourage the proposed Fog computing platform. A fog application developed by Fogos offers distributed processing and computing using fog resources with proximity while taking advantage of location transparency, self-healing, auto-scaling, and mobility support. Our prototype implementation of Fogos shows the benefits of fog computing: fault-tolerant distributed processing over heterogeneous, unreliable, fog nodes while reducing overall latency, thanks to Fogos enabling processing close to the data source.