Machine Learning Based Container Placement in On-Demand Clustered Fogs
Peter Farhat, Sarhad Arisdakessian, Omar Abdul Wahab, Azzam Mourad, Hakima Ould‐Slimane · 2022 International Wireless Communications and Mobile Computing (IWCMC) · 2022
Fog computing extends the concept of cloud computing by allowing services, embedded into virtual machines or containers, to be placed at the edge of the network in the proximity of the end devices. However, due to the huge increase in the number of user requests, placing containers onto fog devices becomes a challenging task. In this work, we address the problem of large-scale container placement in fog computing environments. We propose a machine learning-based K-means clustering solution, which we integrate into the Genetic Algorithm (GA) to improve the selection of the initial population. We first formulate the container placement problem as a multi-objective optimization model with several (conflicting) objectives and then propose a cluster-based GA approach to solve the problem in an efficient manner. Simulation results suggest that our solution outperforms one state-of-the-art approach in terms of effectiveness and efficiency.