Real-Time Task Assignment in Fog/Cloud Network Environments for Profit Maximization

This paper proposes a heuristic model for the real-time task assignment problem in fog/cloud computing networks. Its purpose is to maximize profit by assigning tasks (either to fog nodes or the cloud) while simultaneously considering green computing practices (i.e., changing the states of fog nodes). Compared to the optimal model introduced in [1], the proposed real-time task assignment heuristic performs extremely well considering the fact that it only has access to partial information as opposed to the full historical data for the optimal model. The relative profit returned by the heuristic is consistently within the 79 to 89% range with a processing time that is significantly improved thus, enabling service providers to use such a model in larger network environments and in real-time.