Improving fog computing performance via Fog-2-Fog collaboration

In the Internet of Things (IoT) era, a large volume of data is continuously emitted from a plethora of connected devices. The current network paradigm, which relies on centralised data centres (aka Cloud computing), has become inefficient to respond to IoT latency concern. To address this concern, fog computing allows data processing and storage “close” to IoT devices. However, fog is still not efficient due to spatial and temporal distribution of these devices, which leads to fog nodes’ unbalanced loads. This paper proposes a new fog-2-fog (f2f) collaboration model that promotes offloading incoming requests among fog nodes, according to their load and processing capabilities, via a novel load balancing known as Fog Resource manAgeMEnt Scheme (FRAMES). A formal mathematical model of f2f and FRAMES has been formulated, and a set of experiments has been carried out demonstrating the technical doability of f2f collaboration. The performance of the proposed fog load balancing model is compared to other load balancing models