Optimising QoS-Assurance, Resource Usage and Cost of Fog Application Deployments

Identifying the best application deployment to distribute application components in Fog infrastructures – spanning the IoT-to-Cloud continuum – is a challenging task for application deployers. Indeed, it requires fulfilling all application requirements, whilst determining a trade-off among different objectives (i.e., QoS assurance, Fog resource consumption and cost), resulting in a complex and time-consuming decision-making process to be tuned manually. In this paper, we present a simple multi-objective optimisation scheme that permits selecting the best placement of application components, balancing the trade-off among QoS-assurance, Fog resource consumption and monthly deployment costs. We exploit our prototype, extended with parallel Monte Carlo simulations, and a motivating example to show how IT experts can benefit from our approach.