On the Emerging Coexistence of Edge, Fog and Cloud Computing paradigms in Real-Time Internets-of-EveryThings which operate in the Big-Squared Data space

Abstract

This report motivates the roles of Cloud computing, Edge computing, and the hierarchically distributed cooperative Fog computing, for the real-time analysis of big-data in Internets-of-Everything (IoEs). IoEs are enhanced Internets of Things (IoTs) which integrate people, process, data and heterogeneous “Things”: compute, storage, and sensor/actuator hardware. The ubiquitousness of IoE devices, the ever-increasing amount of big data in IoEs, and the need for real-time computing in IoEs have motivated the problem of distributed data storage and analysis. With trillions (big-data scale) of IoE devices on the verge of being deployed in tomorrow's ever-connected and autonomous society, and with the expected big-data generated by each such IoE device (typically image data of the order of tens and hundreds of gigabytes per day), we are rapidly approaching Big-Squared data dimensions. The power consumption of traditional cloud data centers are already about 70% of all power generated, and it will increase exponentially if Cloud computing is the only solution for tomorrow's IoEs. Moreover, the big-squared data from merging IoEs will create network and compute level bottlenecks that will be impractical from a realtime standpoint, especially in case of rapid mobility in IoEs. Hence, the need for distributed hierarchical Fog computing and associated data management. We survey key features of emerging IoEs, the existing big-data computing and storage frameworks, and point out their capabilities and deficiencies. Finally, we discuss the design and implementation of our Fog computing architecture.